AI timeline

Anthropic Anthropic's S-1 amendment adds Q3 results and $895 billion in commitments

Amendment No. 1 reports a $567 million third-quarter operating loss and about $895.4 billion of compute and infrastructure commitments; the price remains blank.

Source

Markets AI spending worries drive a tech sell-off

The Nasdaq fell 2.9%, its largest one-day percentage drop since June 5, after results from Alphabet, Amazon and Intel raised concern about AI spending.

Source

Mistral Mistral Large 4 weights released

Mistral published the improved Large 4 weights under a modified MIT licence that requires a commercial licence for model-as-a-service businesses.

Source

Canada Canada takes a stake in Cohere

Prime Minister Mark Carney announced an equity investment of up to $1.6 billion in Cohere, tied to jobs, compute in Canada and keeping its headquarters here.

Source

Google Gemini 4 Carbon

Google announced Gemini 4 Carbon, its strongest model, available first to a subset of cyber defenders through the Fairwind Program.

Source

SK hynix SK hynix posts record quarter and expects memory shortage through 2030

SK hynix reported 104.6 trillion won in revenue and said more than half of 2027 and 2028 capacity is under long-term contracts.

Source

OpenAI GPT-6.1 Terra

OpenAI released GPT-6.1 Terra, a fast model for computer use, vision and office work at $1 and $5 per million tokens, rated Critical for cybersecurity.

Source

Anthropic Anthropic files publicly for an IPO

Anthropic filed a Form S-1 with the SEC to list Class A shares on the Nasdaq under the symbol ANT, without a price or share count.

Source

Nvidia Nemotron 3.5 Bolt E4B

NVIDIA released Bolt E4B, post-trained from Google's Gemma 4 E4B, with a 4 GB version for edge devices.

Source

MiniMax MiniMax M3.1-Flash weights

MiniMax published M3.1-Flash with sparse attention throughout and a 252 GB download.

Source

Nvidia NVIDIA's 64 GB DGX Spark goes on sale through partners

A 64 GB DGX Spark goes on sale from Acer, ASUS, Dell, Gigabyte, HP and MSI, starting at US$4,999.

Source

Anthropic Anthropic previews Claude Duet

Anthropic opened a waitlisted US research preview of Claude Duet, a proactive agent built on Claude Opus 5.5 for professionals and small business owners.

Source

DeepSeek DeepSeek-V4.1-Flash-1017

DeepSeek published a minor update to V4.1-Flash reporting 77.0 on DeepSWE v1.1, up from 74.2.

Source

Reflection AI Beam weights released

Reflection published the Beam weights on Hugging Face under Apache 2.0, with over 40 partners and integrations.

Source

Meta Muse Spark Open weights

Meta released Muse Spark Open, 816 billion parameters and 42 billion active, under the Muse Model License.

Source

Nvidia Nemotron 3.5 Super VL and Bolt 74B

NVIDIA released Super VL, a 120 billion parameter vision model.

Source

TSMC TSMC reports $46.72 billion quarter, with 68% from high-performance computing

TSMC reported third-quarter revenue up 41.2% in US dollars.

Source

Intel Intel GPU D-series (Crescent Island)

Intel shared details of its GPU D-series for inference, previously codenamed Crescent Island.

Source

Meta Muse Wonder and Muse Wonder Full

Meta released Muse Wonder, its first frontier-class model, via its API and Muse Code, with Wonder Full for subscribers.

Source

Anthropic Anthropic updates its Usage Policy

Anthropic published a new Usage Policy, effective November 12, banning sustained needless cruelty toward its models and removing its blanket ban on personalized political targeting.

Source

Anthropic Claude Haiku 5.5

Anthropic released Haiku 5.5, a small low-cost model for high-volume tasks, and halved Sonnet 5.5's cache-read price.

Source

OpenAI GPT-6 Instant and Intelligent UI for everyone

OpenAI brought GPT-6 to all ChatGPT tiers with interactive 'Intelligent UI' answers, paid users first and Free and Go users a day later.

Source

Reflection AI Reflection announces Beam

Reflection AI unveiled Beam, its first open-weight model, with 501 billion parameters and 23 billion active.

Source

Aleph Alpha Aleph Alpha releases Kolibri-1

Aleph Alpha released Kolibri, an English-German model of 78 billion parameters, under Apache 2.0.

Source

Cloudflare Cloudflare's Clef decision models

Cloudflare released Clef and Clef-flash, decision models post-trained from Qwen3.8-27B that return typed answers with probabilities, under Apache 2.0.

Source

Moonshot AI Kimi K3.1 Pro

Moonshot launched K3.1 Pro, a further post-trained K3, on its services and promised the weights by November 4.

Source

OpenAI OpenAI launches dots, always-on agents

OpenAI announced dots at DevDay, personal agents powered by GPT-6 Astra, for ChatGPT Pro and Business Premium users in eligible markets.

Source

OpenAI DevDay 2026 and GPT-6.1 Sol

At DevDay in San Francisco OpenAI launched GPT-6.1 Sol, near Astra-level at a fifth of the price, and always-on agents called Dots; GPT-6.1 Astra was held back.

Source

Anthropic Claude Sonnet 5.5

Anthropic released Sonnet 5.5, a faster lower-cost companion to Opus 5.5.

Source

Xiaomi MiMo-V2.6 series

Xiaomi published MiMo-V2.6-Pro and Flash under MIT.

Source

Cohere Cohere and Aleph Alpha sign a definitive merger agreement

The companies signed a definitive agreement, with dual headquarters in Toronto and Berlin, still subject to regulatory approvals.

Source

Bell Canada Bell expands its Saskatchewan AI plan

Saskatchewan and Bell announced a path to 1.2 GW and more than $50 billion in estimated investment.

Source

DeepSeek DeepSeek-V4.1-Flash

DeepSeek released V4.1-Flash, with 552 billion parameters, 8 and 16 billion active, and began phasing out V4-Pro.

Source

Meta Meta launches the Muse personal agent

Meta launched Muse, a personal AI agent that runs on a dedicated cloud virtual machine, in the US, free for most uses with subscription plans.

Source

OpenAI GPT-6 Astra

OpenAI unveiled GPT-6 Astra, a model strong at computer use, for approved testers, with a restricted version for paid ChatGPT users the next day.

Source

U.S. National Science Foundation NSF funds the NAIRR Operations Center

The NSF awarded the San Diego Supercomputer Center a five-year, $35 million award, with TACC as partner, to move NAIRR from pilot to long-term capability.

Source

AI Singapore SEA-LION v4.8

AI Singapore released SEA-LION v4.8, with models built on NVIDIA Nemotron-3 in 30B and 120B sizes.

Source

Z.ai GLM-5.3 weights released

Z.ai published GLM-5.3 under a custom licence.

Source

Alibaba (Qwen) Qwen3.8-Flash-Next

Alibaba published Qwen3.8-Flash-Next, a multimodal mixture-of-experts model of 180 billion parameters, under the Qwen Community License.

Source

Apple Apple announces the Mac Studio with M5 Max and M5 Ultra

The new Mac Studio offers up to 512 GB of unified memory and 1.2 TB/s of memory bandwidth on the M5 Ultra.

Source

DeepSeek DeepSeek-V4-Flash-Vision-Exp

DeepSeek added multimodal input to its API with V4-Flash-Vision-Exp.

Source

Alibaba (Qwen) Qwen3.8-27B

Alibaba released Qwen3.8-27B, a dense vision-language model, under Apache 2.0.

Source

Nvidia Nemotron 3.5 Lightning

NVIDIA released Nemotron 3.5 Lightning, 30 billion parameters and 3 billion active, under OpenMDW 1.1.

Source

Nvidia Nvidia guarantees up to $105 billion for an OpenAI campus

Nvidia agreed to guarantees capped at $105 billion on leases for about 4.25 gigawatts in Ohio, tied to an OpenAI affiliate.

Source

Z.ai GLM-5.3-Flash

Z.ai released GLM-5.3-Flash, a new natively multimodal base of 320 billion parameters and 18 billion active, under MIT.

Source

European Union EuroHPC launches the AI Gigafactories call

EuroHPC opened a tender for up to seven AI gigafactories, with bids due November 12, 2026.

Source

Moonshot AI Kimi K3 weights released

Moonshot published the 1.56 TB Kimi K3 weights on Hugging Face under the custom Kimi K3 License.

Source

Swiss AI (EPFL, ETH Zurich) Apertus 1.5

The Swiss AI Initiative released Apertus 1.5, adding images, optional reasoning and a 262,144-token context window to its fully open 8B and 70B models.

Source

Anthropic Claude Opus 5

Anthropic released Opus 5, close to Fable 5 capability at about half the price.

Source

AMD AMD launches MI455X and Helios

At Advancing AI 2026, AMD said its Helios rack, with 72 MI455X GPUs, was in production, and OpenAI said it expected to start bringing Helios online in the fourth quarter of 2026.

Source

AMD AMD and Anthropic sign a deal for up to 2 GW of MI450 GPUs

Anthropic will deploy up to 2 GW of AMD MI450 GPUs from 2027, and AMD will invest up to $5 billion in Anthropic.

Source

China China-backed WAICO is established

CSIS reports that representatives of 29 countries signed an agreement in Shanghai creating the World Artificial Intelligence Cooperation Organization.

Source

Moonshot AI Kimi K3 launches

Moonshot launched Kimi K3 as a hosted model of 2.8 trillion parameters, promising the weights by July 27.

Source

Thinking Machines Lab Thinking Machines releases Inkling

Thinking Machines Lab released Inkling, its first general model, with 975 billion parameters and 41 billion active, under Apache 2.0.

Source

US states New York imposes the first state data center moratorium

Governor Hochul's Executive Order 62 paused state permitting for data centers of 50 MW or more during an environmental review.

Source

Meta Meta expands its Louisiana site to 5 GW

Meta said its Richland Parish site, home of the Hyperion cluster, will expand to 5 GW with more than $50 billion invested.

Source

OpenAI GPT-5.6 (Luna, Terra, Sol) and ChatGPT Work

OpenAI publicly released the GPT-5.6 family in three tiers and launched ChatGPT Work, an agent combining ChatGPT and Codex for office tasks.

Source

OpenAI Atlas browser shutdown announced

OpenAI said ChatGPT Atlas would close on 9 August 2026, with its features moving into the ChatGPT desktop app and a Chrome extension.

Source

Anthropic AMD and Anthropic sign a chip and stock agreement

AMD committed to buy up to $5 billion of Anthropic stock as GPU milestones are met, and Anthropic expects to spend over $20 billion on AMD systems.

Source

OpenAI GPT-5.6 limited preview

OpenAI gave GPT-5.6 to a limited group and delayed public rollout after a US government request for early access.

Source

China China's LineShine tops the TOP500

China's LineShine took first place on the TOP500 at 2.198 exaflops, which Tom's Hardware reports was reached using only CPUs.

Source

Z.ai GLM-5.2

Z.ai released GLM-5.2 under MIT with a one-million-token context window.

Source

MiniMax MiniMax M3 weights released

MiniMax published M3, 428 billion parameters and 23 billion active, under its Community License.

Source

xAI SpaceX prices a record $75 billion IPO

SpaceX, which owns xAI, priced a $75 billion IPO at $135 a share; with the underwriters’ full option it sold 638.9 million shares for about $85.7 billion net.

Source

Anthropic Claude Fable 5 and Mythos 5

Anthropic released Fable 5, a public Mythos-class model with safeguards, and a limited-access Mythos 5.

Source

Apple Apple introduces its third generation of Foundation Models

Apple's lineup includes AFM 3 Core at 3 billion parameters and AFM 3 Core Advanced at 20 billion, stored in flash with 1 to 4 billion active.

Source

Google Alphabet raises about $50 billion in stock and related securities

Alphabet raised about $50 billion net from a public share sale, a Berkshire Hathaway placement and convertible preferred stock, partly for AI infrastructure.

Source

US states New York legislature passes a data center pause

The legislature passed a one-year moratorium on state permits for data centers of 20 MW or more.

Source

Canada Canada's national AI strategy, AI for All

Canada launched AI for All, a national AI strategy built on a build-partner-buy approach to sovereign compute, models and adoption.

Source

European Union Commission proposes the Chips Act 2.0

The European Commission proposed a Chips Act 2.0 as part of a technological sovereignty package.

Source

Nvidia Vera Rubin enters full production

Nvidia said Vera Rubin was ramping into full production, with production shipments due to start that fall.

Source

Anthropic Anthropic raises $65 billion at $965 billion

Anthropic announced a $65 billion Series H at a $965 billion post-money valuation, including $15 billion of earlier commitments.

Source

Anthropic Claude Opus 4.8

Anthropic released Claude Opus 4.8.

Source

xAI SpaceX, owner of xAI, files its S-1

SpaceX publicly filed to list on Nasdaq, disclosing xAI segment results and $12.7 billion of 2025 AI capital spending.

Source

Google Gemini 3.5 Flash

Google launched Gemini 3.5 Flash, positioned around agents.

Source

Cerebras Cerebras begins trading on Nasdaq

Cerebras Class A shares began trading on the Nasdaq under the ticker CBRS, after pricing at 185 US dollars.

Source

OpenAI GPT-5.5 with trusted access for cyber

OpenAI announced a GPT-5.5 variant for vetted cybersecurity defenders.

Source

OpenAI GPT-5.5 Instant becomes default

GPT-5.5 Instant replaced GPT-5.3 Instant as ChatGPT's default model.

Source

Cohere Cohere to merge with Aleph Alpha

Canada's Cohere announced it would combine with Germany's Aleph Alpha, backed by Schwarz Group, to offer a sovereign alternative to American AI providers.

Source

OpenAI GPT-5.5 (codename Spud)

OpenAI released GPT-5.5 Thinking and Pro, with API access the next day.

Source

Google Google announces eighth-generation TPU 8t and TPU 8i

Google announced its eighth-generation TPUs as two specialized chips, the TPU 8t for training and the TPU 8i for inference.

Source

Anthropic Claude Design preview

Anthropic previewed Claude Design, which turns prompts into visual prototypes, for paid users.

Source

Anthropic Claude Opus 4.7

Anthropic released Claude Opus 4.7.

Source

xAI NAACP sues xAI over Southaven turbines

The NAACP sued xAI and MZX Tech under the Clean Air Act over gas turbines allegedly run without a permit.

Source

Meta Muse Spark

Meta Superintelligence Labs released Muse Spark, its first model, to replace Llama in Meta's assistant.

Source

Anthropic Claude Mythos Preview (Project Glasswing)

Anthropic gave a limited Mythos Preview model to about a dozen organizations for defensive cybersecurity work.

Source

Broadcom Broadcom discloses long-term TPU deal with Google and 3.5 GW for Anthropic

Broadcom's filing describes custom TPU supply for Google and about 3.5 GW of capacity for Anthropic from 2027.

Source

Google Gemma 4 arrives in the Android AICore Developer Preview

Google offers Gemma 4 E2B and E4B on Android, and says Gemma 4 is the foundation for the next Gemini Nano.

Source

US government American AI Exports Program opens its first call

Commerce opened proposals for full-stack AI export packages, through June 30, 2026.

Source

OpenAI OpenAI closes $122 billion at an $852 billion valuation

OpenAI closed a round of $122 billion in committed capital, including up to $50 billion from Amazon and $30 billion each from Nvidia and SoftBank.

Source

Meta Entergy and Meta sign Louisiana power agreement

Entergy Louisiana announced seven new gas plants above 5,200 MW for Meta's data center, with about $2 billion in projected customer savings.

Source

OpenAI OpenAI announces end of Sora app and API

OpenAI said it would discontinue the Sora consumer app and API to free compute for other products.

Source

Blue Origin Blue Origin files for Project Sunrise

Blue Origin filed with the FCC for a space-based computing network of more than 50,000 satellites.

Source

OpenAI GPT-5.4 mini and nano

OpenAI released GPT-5.4 mini, available to free users, and GPT-5.4 nano for the API only.

Source

Nvidia Vera Rubin at GTC 2026, with the Groq 3 LPX rack

At GTC, Nvidia presented the Vera Rubin platform, including the Vera Rubin NVL72 GPU rack and the Groq 3 LPX inference accelerator rack.

Source

Micron Micron in high-volume HBM4 production

Micron said its 36 GB 12-high HBM4, designed for Nvidia's Vera Rubin, was in high-volume production at over 2.8 TB/s.

Source

Meta Meta four-chip MTIA roadmap

Meta said it was developing four MTIA generations, with MTIA 300 already in production for ranking and recommendation training.

Source

OpenAI GPT-5.4 Thinking and Pro

OpenAI released GPT-5.4 in Thinking and Pro versions for paid ChatGPT users and the API.

Source

Apple Apple M5 Pro and M5 Max

Apple debuted the M5 Pro and M5 Max, with unified memory bandwidth up to 614 GB per second.

Source

OpenAI GPT-5.3 Instant

OpenAI made GPT-5.3 Instant the ChatGPT default, aiming for fewer needless refusals and better web answers.

Source

Amazon Web Services Drone strikes hit AWS data centres in the Gulf

Drones struck two Amazon data centres in the UAE and a nearby strike damaged one in Bahrain during the Iran war.

Source

OpenAI OpenAI and Oracle drop the Abilene expansion

OpenAI reversed its plan to expand Abilene to 2.1 GW, and Microsoft partnered with Crusoe on the adjacent 900 MW.

Source

Canada Hydro-Québec proposes a 13-cent data center rate

Hydro-Québec asked the Régie de l'énergie to approve an average rate of 13 cents/kWh for data centers above 5 MW.

Source

Google Gemini 3.1 Pro

Google released Gemini 3.1 Pro.

Source

Anthropic Claude Sonnet 4.6

Anthropic released Sonnet 4.6, a faster and cheaper model.

Source

Alibaba (Qwen) Qwen3.5

Alibaba released Qwen3.5.

Source

OpenAI GPT-4o retired from ChatGPT

GPT-4o and several older models were removed from ChatGPT; business plans kept GPT-4o in custom GPTs until 3 April 2026.

Source

Samsung Samsung ships commercial HBM4

Samsung said it had started shipping commercial HBM4, at 11.7 gigabits per second.

Source

Z.ai GLM-5

Z.ai released GLM-5 under MIT, growing to 744 billion parameters with 40 billion active.

Source

Anthropic Anthropic raises $30 billion at $380 billion

Anthropic announced a $30 billion Series G led by GIC and Coatue at a $380 billion post-money valuation.

Source

Anthropic Claude Opus 4.6

Anthropic released Opus 4.6 with agent teams and Claude in PowerPoint.

Source

OpenAI GPT-5.3-Codex

OpenAI released GPT-5.3-Codex, a faster coding model pitched for broader computer work beyond code.

Source

xAI xAI becomes part of SpaceX

SpaceX absorbed xAI, making it a wholly owned subsidiary, according to its later IPO filing.

Source

OpenAI Codex app for macOS

OpenAI released a desktop Codex app for managing several coding agents in parallel, with automations; Windows followed on 4 March.

Source

US Department of Commerce US grants Nvidia a first H200 licence for China

Nvidia's annual report says the US government granted a February 2026 licence to ship small amounts of H200 chips to specific China-based customers, with US inspection and a 25% tariff.

Source

Government of India India adds GPUs at the AI Impact Summit

At the AI Impact Summit, India announced 20,000 more GPUs on top of 38,000 provisioned.

Source

Canada BC launches competitive call for data center power

BC Hydro and the province launched a call under limits of 100 MW for conventional and 300 MW for AI data centers over two years.

Source

OpenAI Retirement of GPT-4o in ChatGPT announced

OpenAI said GPT-4o, GPT-4.1, GPT-4.1 mini and o4-mini would leave ChatGPT on 13 February 2026, with the API unchanged.

Source

Microsoft Microsoft Maia 200

Microsoft announced Maia 200, an inference accelerator on TSMC's 3-nanometre process with 216 GB of HBM3e memory.

Source

European Union EU Council clears AI gigafactories

The Council extended EuroHPC's mandate to cover AI gigafactories.

Source

OpenAI ChatGPT Go worldwide; ad tests announced

OpenAI made the $8 ChatGPT Go plan available everywhere and said it would test labeled ads for US Free and Go users.

Source

US Department of Commerce US rule moves H200-class chips to case-by-case review

A Commerce rule moved exports of Nvidia H200-class chips to China and Macau from a presumption of denial to case-by-case review.

Source

OpenAI OpenAI adds 750 MW of Cerebras compute

OpenAI said it would add 750 megawatts of low-latency Cerebras compute to its platform.

Source

US government Trump sets a 25% tariff on certain advanced chips

A proclamation placed a 25% tariff on a narrow group of advanced chips, with exemptions such as use in US data centres.

Source

US Bureau of Industry and Security US moves H200-class chips to case-by-case review for China

BIS changed licence review for chips such as the Nvidia H200 and AMD MI325X from presumption of denial to case-by-case, under conditions.

Source

Anthropic Claude Cowork research preview

Anthropic previewed Cowork, an agent in the Claude desktop app that works on local files for non-programmers, starting with Max subscribers.

Source

Apple Apple and Google Gemini partnership

Apple and Google said the next Apple Foundation Models would be built on Gemini models and cloud technology.

Source

AMD AMD previews Helios at CES

At CES, AMD showed its Helios rack-scale platform and the Instinct MI455X accelerator.

Source

Nvidia Nvidia details the Rubin platform

At CES, Nvidia detailed Rubin as six new chips forming one AI supercomputer, with partner products due in the second half of 2026.

Source

Moonshot AI Kimi K2.5

Moonshot added native vision and agent swarms of up to 100 sub-agents to the K2 line in Kimi K2.5.

Source

SpaceX SpaceX applies for up to one million orbital data center satellites

SpaceX applied to the US FCC for up to one million satellites intended as a distributed data center.

Source

Groq Groq licenses inference technology to Nvidia

Groq announced a non-exclusive licensing agreement giving Nvidia access to its inference technology.

Source

OpenAI GPT-5.2-Codex

OpenAI released a GPT-5.2 version optimized for agentic coding in Codex, with API access added in January 2026.

Source

Google Gemini 3 Flash

Google released Gemini 3 Flash, a faster lower-cost Gemini 3 model.

Source

OpenAI New ChatGPT Images (GPT Image 1.5)

OpenAI rolled out a new image model to all ChatGPT users and the API, with more precise edits and faster generation.

Source

OpenAI GPT-5.2

OpenAI released GPT-5.2 in Instant, Thinking and Pro versions, focused on professional work and long tasks.

Source

Micron Micron announces exit from its Crucial consumer business

Micron said it would leave the Crucial consumer brand to better supply larger strategic customers amid AI demand.

Source

Amazon Web Services AWS Trainium3 and Trn3 UltraServers

AWS launched Trainium3, its first 3-nanometre AI chip, in UltraServers of up to 144 chips, and said it was already working on Trainium4.

Source

Mistral Mistral 3 and Mistral Large 3

Mistral released the Mistral 3 family, including Mistral Large 3 and small Ministral 3 models, under open licences.

Source

DeepSeek DeepSeek-V3.2

DeepSeek released V3.2 with sparse attention for cheaper long-context use.

Source

US government Trump signs the Genesis Mission order

An executive order launched the Genesis Mission, a Department of Energy effort to apply national computing and data to AI for science.

Source

Anthropic Claude Opus 4.5

Anthropic released Opus 4.5 with gains in coding and office work at lower prices.

Source

OpenAI Shopping research in ChatGPT

ChatGPT gained a shopping research mode that builds buyer's guides using a model based on GPT-5 thinking mini.

Source

Google Nano Banana Pro (Gemini 3 Pro Image)

Google released an image model built on Gemini 3 Pro with better text rendering.

Source

OpenAI GPT-5.1-Codex-Max

OpenAI released a coding model for long-running work that compacts its own context to keep going across sessions.

Source

Anthropic Anthropic, Microsoft and Nvidia partnership

Anthropic committed to buy 30 billion US dollars of Azure compute and up to one gigawatt of additional capacity, in a partnership with Microsoft and Nvidia.

Source

EuroHPC JU Contract signed for Alice Recoque

EuroHPC and Eviden signed the contract for Alice Recoque, Europe's second exascale system, to be hosted in France.

Source

Anthropic Microsoft and Nvidia commit to invest in Anthropic

Anthropic committed to $30 billion of Azure compute while Nvidia and Microsoft committed up to $10 billion and $5 billion of investment.

Source

Google Gemini 3 Pro

Google released Gemini 3 Pro in the Gemini app, Search and developer tools.

Source

OpenAI Group chats pilot

OpenAI piloted group chats where up to 20 people can share one ChatGPT conversation, starting in a few countries.

Source

OpenAI GPT-5.1 Instant and Thinking

OpenAI updated ChatGPT to GPT-5.1, aiming for warmer conversation, adaptive reasoning and new personality presets.

Source

US Department of Commerce US suspends the affiliates rule for a year

The US suspended its rule extending export limits to affiliates of listed entities, until November 9, 2026.

Source

China China pauses its October export controls

China's commerce ministry suspended the six October 9 announcements for one year.

Source

Google Google says its Ironwood TPU will be generally available within weeks

Google announced that Ironwood, its seventh-generation TPU, would be generally available in the coming weeks, alongside new Axion virtual machines.

Source

Google Google unveils Project Suncatcher

Google Research published a design for solar-powered satellite constellations carrying TPUs, with two prototypes planned with Planet by early 2027.

Source

Canada Alberta allocates its 1,200 MW data center limit

Alberta's grid operator allocated its interim 1,200 MW limit to two projects, leaving 37 proposals seeking 19.4 GW.

Source

US Department of Energy US announces Solstice and Equinox at Argonne

The DOE, Nvidia and Oracle announced Solstice, with 100,000 Blackwell GPUs, and Equinox, with 10,000, at Argonne.

Source

Qualcomm Qualcomm AI200 and AI250

Qualcomm announced AI200 and AI250 inference accelerators and racks for data centres, due in 2026 and 2027.

Source

Anthropic Anthropic plans up to one million Google TPUs

Anthropic said it would expand its use of Google Cloud with up to one million TPUs, in a deal worth tens of billions of dollars.

Source

Meta Meta and Blue Owl form a $27 billion Hyperion venture

Meta kept 20% of a joint venture that will own its Louisiana campus, with Blue Owl funds holding 80% and costs of about $27 billion.

Source

OpenAI ChatGPT Atlas browser

OpenAI launched ChatGPT Atlas, a macOS web browser with ChatGPT built in and an agent mode for paid users.

Source

Apple Apple M5

Apple announced the M5, with 153 GB per second of unified memory bandwidth.

Source

Anthropic Claude Haiku 4.5

Anthropic released Haiku 4.5, a small fast model.

Source

Intel Intel announces the Crescent Island inference GPU

Intel announced a data-centre GPU for inference, using Xe3P and 160 GB of LPDDR5X, with customer samples expected in the second half of 2026.

Source

OpenAI OpenAI and Broadcom custom accelerators

OpenAI and Broadcom announced a collaboration to deploy 10 gigawatts of OpenAI-designed accelerators.

Source

Nvidia Nvidia DGX Spark ships

Nvidia said it and its partners were shipping DGX Spark that week, with 128 GB of unified memory and a claimed petaflop of AI performance, and that orders would open on October 15.

Source

Microsoft Microsoft's first large GB300 cluster

Microsoft Azure said it had delivered the first at-scale production cluster with more than 4,600 GB300 NVL72 systems, for OpenAI workloads.

Source

China China widens rare earth export controls

China issued six announcements widening export controls, including on rare earths.

Source

OpenAI AMD gives OpenAI a warrant in a 6-gigawatt chip deal

AMD issued OpenAI a warrant for up to 160 million shares that vests as OpenAI deploys up to 6 gigawatts of AMD GPUs.

Source

AMD AMD and OpenAI sign a 6 GW GPU agreement

OpenAI agreed to deploy up to 6 GW of AMD GPUs and received a warrant for up to 160 million AMD shares.

Source

OpenAI OpenAI DevDay 2025

OpenAI announced apps inside ChatGPT with the Apps SDK, AgentKit, general availability of Codex, and Sora 2 and GPT-5 Pro in the API.

Source

OpenAI Apps in ChatGPT and the Apps SDK

OpenAI let third-party apps run inside ChatGPT conversations, built on the Model Context Protocol.

Source

OpenAI AgentKit and Agent Builder

OpenAI released AgentKit, including the visual Agent Builder, a connector registry and ChatKit for embedding agents.

Source

TSMC TSMC starts 2 nm volume production

TSMC says its 2-nanometre process, its first with nanosheet transistors, started volume production in the fourth quarter of 2025.

Source

Netherlands Netherlands invokes emergency powers over Nexperia

The Dutch government invoked the Goods Availability Act over the Chinese-owned chipmaker Nexperia, over fears of technology transfer.

Source

OpenAI Sora 2 and the Sora app

OpenAI released Sora 2, a video model with synchronized audio, alongside an invite-only social iOS app with cameos.

Source

US Department of Commerce US extends Entity List to affiliates

The US Commerce Department extended end-user controls to affiliates of listed entities.

Source

DeepSeek DeepSeek-V3.2-Exp

DeepSeek released the experimental V3.2.

Source

Anthropic Claude Sonnet 4.5

Anthropic released Sonnet 4.5, focused on coding and long-running agents, with the Claude Agent SDK.

Source

Telus Telus opens its Rimouski AI factory

Telus opened a sovereign AI factory in Rimouski, Quebec, using Nvidia GPUs and HPE infrastructure.

Source

OpenAI Stargate adds five sites

OpenAI, Oracle and SoftBank announced five new Stargate sites, bringing planned capacity to nearly 7 gigawatts.

Source

Nvidia OpenAI and Nvidia 10 GW partnership

OpenAI and Nvidia announced a letter of intent to deploy at least 10 gigawatts of Nvidia systems.

Source

OpenAI Nvidia and OpenAI announce a 10-gigawatt letter of intent

Nvidia said it intends to invest up to $100 billion in OpenAI as each gigawatt of its systems is deployed.

Source

Microsoft Microsoft Fairwater

Microsoft introduced Fairwater in Wisconsin, which it called its largest AI data centre.

Source

Nvidia Nvidia and Intel partner on data centre and PC chips

Nvidia said it would invest 5 billion US dollars in Intel, and the two agreed to develop custom data centre and PC products together.

Source

Huawei Huawei publishes its Ascend roadmap to 2028

Eric Xu set out Ascend 950PR for early 2026, 950DT for late 2026, 960 for late 2027 and 970 for late 2028, with Huawei's own high-bandwidth memory.

Source

Simon Fraser University SFU brings Fir online

Simon Fraser University announced Fir, replacing Cedar after about $80 million in upgrades, as Canada's most powerful academic supercomputer.

Source

OpenAI GPT-5-Codex

OpenAI released GPT-5-Codex, a GPT-5 version tuned for agentic coding, as the default for Codex cloud tasks and code review.

Source

SK hynix SK hynix completes HBM4 development

SK hynix said it had completed HBM4 development and was ready for mass production, with speeds above 10 gigabits per second.

Source

Mistral Mistral AI raises €1.7 billion

Mistral AI announced a €1.7 billion Series C at an €11.7 billion post-money valuation.

Source

EuroHPC JU JUPITER inaugurated at Jülich

Germany's chancellor and an EU commissioner inaugurated JUPITER, Europe's first exascale machine, with about 24,000 Nvidia GH200 chips.

Source

Swiss AI (EPFL, ETH Zurich) Apertus, Switzerland's public open model

EPFL, ETH Zurich and CSCS released Apertus, a fully open multilingual model (8B and 70B) with public weights, data recipes and code, presented as public AI.

Source

Z.ai GLM-4.6

Z.ai released GLM-4.6 under MIT.

Source

OpenAI gpt-realtime and Realtime API GA

OpenAI made the Realtime API generally available and released gpt-realtime, a speech-to-speech model with SIP calling and MCP support.

Source

Google Gemini 2.5 Flash Image (Nano Banana)

Google released an image editing and generation model, nicknamed Nano Banana, in the Gemini app.

Source

US government US takes an equity stake in Intel

Intel said the US government would invest $8.9 billion in its common stock, funded largely by CHIPS Act grants not yet paid.

Source

DeepSeek DeepSeek-V3.1

DeepSeek released V3.1 with hybrid thinking and non-thinking modes.

Source

Z.ai GLM-4.5V

Z.ai released GLM-4.5V, a vision-language model of 108 billion parameters, under MIT.

Source

Tesla Tesla disbands Dojo team

Bloomberg reported that Tesla was disbanding its Dojo supercomputer team and that its leader, Peter Bannon, was leaving.

Source

SciNet SciNet launches the Trillium supercomputer in Toronto

The University of Toronto's SciNet launched Trillium, with 1,224 CPU nodes and a GPU section, replacing Niagara and Mist.

Source

OpenAI GPT-5 released

OpenAI launched GPT-5, a unified system that routes between fast and thinking models, to all ChatGPT users including free.

Source

Anthropic Claude Opus 4.1

Anthropic released Opus 4.1 with improved coding and agentic performance.

Source

OpenAI gpt-oss open-weight models

OpenAI released gpt-oss-120b and gpt-oss-20b, its first open-weight language models since GPT-2, under Apache 2.0.

Source

Broadcom Broadcom ships Jericho4

Broadcom shipped a router designed to link over one million accelerators across several data centres.

Source

US government U.S. order on exporting the American AI stack

A U.S. executive order set up a program to promote full-stack American AI export packages.

Source

OpenAI OpenAI and Oracle add 4.5 GW to Stargate

OpenAI and Oracle agreed to develop 4.5 gigawatts of additional Stargate capacity in the United States.

Source

United Kingdom UK launches Isambard-AI

The UK's most powerful supercomputer, Isambard-AI, launched in Bristol.

Source

OpenAI ChatGPT agent

OpenAI merged Operator and deep research into ChatGPT agent, which uses its own virtual computer to complete tasks.

Source

Broadcom Broadcom ships Tomahawk Ultra

Broadcom shipped an Ethernet switch aimed at AI scale-up, with 250 ns latency at 51.2 Tb/s.

Source

Nvidia Nvidia says H20 sales to China will resume

Nvidia said it was filing for licences to sell the H20 again and that the US government had assured it licences would be granted.

Source

Malaysia Malaysia requires permits for US AI chips

Malaysia's trade ministry said exports, transhipments and transits of high-performance US-origin AI chips needed a Strategic Trade Permit, effective at once.

Source

Meta Zuckerberg names Meta's Prometheus and Hyperion clusters

Mark Zuckerberg said Meta was building Prometheus, due online in 2026, and Hyperion, able to scale to 5 gigawatts over several years.

Source

Moonshot AI Kimi K2 open weights

Moonshot released Kimi K2, a mixture-of-experts model with 1 trillion parameters and 32 billion active, under a Modified MIT licence.

Source

xAI Grok 4 and Grok 4 Heavy

xAI released Grok 4 and a multi-agent Grok 4 Heavy tier.

Source

Hugging Face SmolLM3

Hugging Face publishes a 3-billion-parameter open model trained on about 11 trillion tokens with 384 H100 GPUs over 24 days, and details its training recipe.

Source

Z.ai GLM-4.5

Z.ai released GLM-4.5 under MIT, with 355 billion total parameters by its model card.

Source

Calcul Québec Calcul Québec launches Rorqual

Calcul Québec launched Rorqual, a research supercomputer at the McGill data centre, ranked 192nd on the June 2025 TOP500 for its GPU part.

Source

Midjourney Midjourney V1 video model

Midjourney released its first video model, animating images into short clips.

Source

Taiwan Taiwan puts Huawei and SMIC on its export control list

Taiwan's Commerce Ministry added Huawei and SMIC to its export control list of strategic high-tech commodities.

Source

AMD AMD Instinct MI350 series

At Advancing AI 2025, AMD launched the MI350 series and laid out a rack-scale roadmap.

Source

Ultra Ethernet Consortium Ultra Ethernet 1.0 specification

The Ultra Ethernet Consortium published its 1.0 specification, an Ethernet-based stack for AI and high-performance computing.

Source

Mistral Magistral reasoning models

Mistral released Magistral, its first reasoning models.

Source

EleutherAI The Common Pile v0.1 and Comma v0.1 models

Researchers release an 8 TB dataset of public domain and openly licensed text, with two 7-billion-parameter models trained on it.

Source

Broadcom Broadcom ships Tomahawk 6 at 102.4 Tb/s

Broadcom began shipping a switch series with 102.4 Tb/s of capacity, which it called the world's first.

Source

OpenAI First GB200 racks reach Stargate Abilene

OpenAI said Oracle began delivering the first Nvidia GB200 racks to Abilene in June 2025.

Source

MiniMax MiniMax-M1

MiniMax released M1 under Apache 2.0 with about 456 billion parameters.

Source

US Department of Energy US names the Doudna supercomputer

The US energy secretary announced a Dell contract for NERSC-10, named Doudna and due in 2026, using Nvidia technology.

Source

Bell Canada Bell launches AI Fabric

Bell announced a plan for upwards of 500 MW of AI compute across six British Columbia facilities.

Source

DeepSeek DeepSeek-R1-0528

DeepSeek released an updated R1.

Source

G42 Stargate UAE announced

G42, OpenAI, Oracle, Nvidia, SoftBank Group and Cisco announced a 1-gigawatt cluster within a 5-gigawatt UAE-U.S. campus.

Source

Anthropic Claude Opus 4 and Sonnet 4

Anthropic released Claude Opus 4 and Sonnet 4 and made Claude Code generally available.

Source

Intel Intel launches the Arc Pro B60 with 24 GB

Intel launched Arc Pro B60 and B50 workstation GPUs, with up to 24 GB per card and AI inference as a stated use.

Source

Nvidia Nvidia NVLink Fusion

Nvidia opened NVLink to partners with NVLink Fusion, a way for other companies' custom chips to join its rack-scale systems.

Source

OpenAI Codex cloud coding agent

OpenAI previewed Codex, a cloud software engineering agent powered by codex-1, a version of o3, for Pro, Team and Enterprise users.

Source

US Department of Commerce US says it will rescind AI Diffusion rule

The US Commerce Department announced it would rescind the AI Diffusion rule.

Source

US government U.S. moves to rescind the AI Diffusion Rule

The Commerce Department announced the rescission of the Biden-era AI Diffusion Rule.

Source

US Bureau of Industry and Security US Commerce announces rescission of the AI Diffusion rule

Commerce said it would rescind the January 2025 three-tier AI Diffusion rule and issued guidance on advanced chips.

Source

Humain Saudi Arabia launches Humain

Saudi Arabia's Crown Prince launched Humain, a Public Investment Fund company spanning the AI value chain.

Source

Mistral Mistral Medium 3

Mistral released Mistral Medium 3, a mid-size model aimed at enterprises.

Source

Xiaomi Xiaomi MiMo-7B

Xiaomi released MiMo-7B, a 7 billion parameter reasoning model, under MIT.

Source

OpenAI GPT-4o update rolled back for sycophancy

OpenAI reverted a GPT-4o update that made ChatGPT overly flattering and explained what went wrong.

Source

Alibaba (Qwen) Qwen3

Alibaba released the Qwen3 family of open-weight models with hybrid thinking.

Source

JEDEC JEDEC releases the HBM4 standard

JEDEC released JESD270-4, the HBM4 standard.

Source

OpenAI OpenAI o3 and o4-mini

OpenAI released o3 and o4-mini, reasoning models that can use all ChatGPT tools and reason with images.

Source

OpenAI Codex CLI

OpenAI released Codex CLI, an open-source coding agent that runs in the terminal.

Source

Telus Telus plans sovereign AI factories

Telus announced plans for AI factories in Rimouski and Kamloops.

Source

Nvidia US requires licence for Nvidia H20 to China

Nvidia disclosed that a licence requirement for H20 exports to China would stay in effect indefinitely, and expected charges of up to about 5.5 billion US dollars.

Source

OpenAI GPT-4.1 prompting guide

OpenAI published a cookbook guide explaining how to prompt GPT-4.1, which follows instructions more literally.

Source

OpenAI GPT-4.1, 4.1 mini and 4.1 nano in the API

OpenAI released three API models with up to one million tokens of context and better coding and instruction following.

Source

Google Google Ironwood, seventh-generation TPU

Google announced Ironwood, its first TPU built for inference, scaling to 9,216 chips and a claimed 42.5 exaflops.

Source

UALink Consortium UALink 200G 1.0 specification

The UALink Consortium released the Ultra Accelerator Link 200G 1.0 specification.

Source

Meta Llama 4 Scout and Maverick

Meta released Llama 4 Scout and Maverick, mixture-of-experts multimodal models, and previewed Behemoth.

Source

Midjourney Midjourney V7

Midjourney released V7 in alpha.

Source

CoreWeave CoreWeave prices its IPO at $40 a share

CoreWeave priced its IPO at $40 a share for 37.5 million shares, about $1.5 billion.

Source

DeepSeek DeepSeek-V3-0324

DeepSeek released an updated V3 model.

Source

Google Gemini 2.5 Pro

Google released Gemini 2.5 Pro, a thinking model that topped several leaderboards.

Source

OpenAI Native image generation in GPT-4o

ChatGPT replaced DALL-E with GPT-4o's built-in image generation, which handles text in images and edits well.

Source

SoftBank SoftBank to buy Ampere Computing

SoftBank agreed to buy chip designer Ampere Computing in an all-cash deal valued at 6.5 billion US dollars.

Source

Nvidia Nvidia co-packaged optics switches

Nvidia announced Spectrum-X and Quantum-X Photonics networking switches using co-packaged optics.

Source

Nvidia Nvidia RTX PRO Blackwell workstation cards

Nvidia brought Blackwell to workstations and servers with its RTX PRO line.

Source

Nvidia Nvidia DGX Spark and DGX Station announced

Nvidia announced DGX Spark, the new name for Project Digits, and DGX Station as personal AI computers.

Source

Nvidia Nvidia Blackwell Ultra

Nvidia announced Blackwell Ultra, including the GB300 NVL72 rack and the HGX B300 NVL16 system.

Source

OpenAI Responses API and Agents SDK

OpenAI launched the Responses API, built-in web search, file search and computer use tools, and the open-source Agents SDK.

Source

Apple Apple M3 Ultra with 512 GB

Apple revealed the M3 Ultra, configurable with up to 512 GB of unified memory.

Source

TSMC TSMC raises US investment to 165 billion dollars

TSMC said it intended to expand its total United States investment to 165 billion US dollars.

Source

OpenAI GPT-4.5 research preview

OpenAI released GPT-4.5, its largest non-reasoning chat model, to Pro users and developers.

Source

Anthropic Claude 3.7 Sonnet and Claude Code preview

Anthropic released Claude 3.7 Sonnet with optional extended thinking and previewed Claude Code, a terminal coding agent.

Source

xAI Grok 3

xAI released Grok 3 with a Think reasoning mode.

Source

OpenAI Deep research in ChatGPT

OpenAI launched deep research, an agent powered by a version of o3 that compiles cited reports from many web sources, for Pro users first.

Source

OpenAI OpenAI o3-mini

OpenAI released o3-mini, a low-cost reasoning model in ChatGPT (including free users) and the API.

Source

Alibaba (Qwen) Qwen2.5-Max

Alibaba released Qwen2.5-Max, a large mixture-of-experts model.

Source

DeepSeek DeepSeek app tops charts; Nvidia shares fall

DeepSeek's app reached the top of the US App Store and AI chip stocks dropped sharply.

Source

OpenAI Operator research preview

OpenAI released Operator, an agent that uses its own browser to complete web tasks, for US Pro users.

Source

OpenAI Stargate Project announced

OpenAI, SoftBank, Oracle and MGX announced a joint venture to build AI data centers in the US.

Source

DeepSeek DeepSeek-R1

DeepSeek released R1, an open-weight reasoning model rivaling o1, under the MIT licence.

Source

US Department of Commerce US AI Diffusion rule

The US published the Framework for Artificial Intelligence Diffusion, setting global licence rules for advanced AI chips and some model weights.

Source

Netherlands Netherlands to license chip measuring tools

The Dutch government said that from April 1, 2025, more technology would need a licence, such as specific measuring and inspection equipment.

Source

US Department of Commerce US tightens foundry checks on AI chips

The US issued rules requiring chip factories to do more checks on customers, to stop advanced chips reaching China.

Source

AMD AMD Ryzen AI Max (Strix Halo)

At CES, AMD announced Ryzen AI Max processors with up to 128 GB of unified memory, of which up to 96 GB can serve graphics.

Source

Nvidia Nvidia Project Digits desktop

Nvidia announced Project Digits, a desktop AI computer built on Grace Blackwell, priced from 3,000 US dollars.

Source

Nvidia GeForce RTX 50 series (Blackwell)

At CES, Nvidia announced the Blackwell-based GeForce RTX 50 series, led by the RTX 5090 with 92 billion transistors.

Source

DeepSeek DeepSeek-V3

DeepSeek released V3, a 671B mixture-of-experts open model reportedly trained at low cost.

Source

OpenAI o3 and o3-mini previewed

On the last day of its '12 days' event OpenAI previewed o3 and o3-mini for safety testing, not public release.

Source

OpenAI o1 in the API

OpenAI brought o1 to the API with function calling and structured outputs, and updated the Realtime API.

Source

Google Google's Trillium TPU reaches general availability

Google made its sixth-generation TPU generally available and said it trained Gemini 2.0 on it.

Source

Google Gemini 2.0 Flash experimental

Google introduced Gemini 2.0, starting with an experimental Flash model aimed at agentic uses.

Source

European Union EU selects its first seven AI Factories

EuroHPC selected seven AI Factories in Finland, Germany, Greece, Italy, Luxembourg, Spain and Sweden.

Source

Microsoft Microsoft announces zero-water-for-cooling data center design

Microsoft said all new data center designs from August 2024 use closed-loop chip-level cooling that avoids evaporating water.

Source

OpenAI Sora released to the public

OpenAI launched Sora Turbo video generation for ChatGPT Plus and Pro users at sora.com.

Source

Canada Canadian Sovereign AI Compute Strategy

Canada launched a strategy investing up to $2 billion in AI computing: data centres, a public supercomputer and help for small businesses to pay for compute.

Source

OpenAI Full o1 and ChatGPT Pro

OpenAI released the full o1 model and a $200 monthly Pro plan with o1 pro mode.

Source

Meta Louisiana announces Meta's $10 billion Richland Parish data centre

Louisiana's governor announced that Meta chose Richland Parish for a $10 billion AI data centre.

Source

Amazon Web Services AWS Trn2 instances generally available

AWS made EC2 Trn2 instances generally available.

Source

China China bans gallium and germanium exports to the US

China's commerce ministry barred exports of gallium, germanium, antimony and superhard materials to the United States.

Source

US Department of Commerce US controls on HBM and new foreign direct product rules

The US Commerce Department added controls on High Bandwidth Memory and new foreign direct product rules for chipmaking equipment.

Source

Meta Meta breaks ground in Richland Parish, Louisiana

Meta broke ground on its Richland Parish data center.

Source

Anthropic Model Context Protocol (MCP)

Anthropic open-sourced MCP, a standard for connecting AI assistants to data sources and tools.

Source

US Department of Energy El Capitan tops the TOP500

Lawrence Livermore's El Capitan, on AMD MI300A accelerators, took first place on the TOP500 with 1.742 exaflops.

Source

OpenAI ChatGPT search

OpenAI built web search with source links into ChatGPT, starting with paid users.

Source

UALink Consortium UALink Consortium incorporates

The Ultra Accelerator Link group incorporated and opened membership.

Source

Apple First Apple Intelligence features ship

Apple released iOS 18.1 with the first Apple Intelligence features in US English.

Source

Anthropic Computer use beta, new 3.5 Sonnet and 3.5 Haiku

Anthropic let Claude control a computer through screenshots, mouse and keyboard in public beta, alongside upgraded models.

Source

AMD AMD Instinct MI325X

AMD announced the MI325X accelerator with 256 GB of HBM3E memory.

Source

OpenAI Canvas in ChatGPT

OpenAI introduced canvas, a side-by-side workspace for writing and coding with ChatGPT.

Source

OpenAI DevDay 2024: Realtime API

OpenAI announced the Realtime API for speech-to-speech apps, plus prompt caching, vision fine-tuning and model distillation.

Source

SK hynix SK hynix 12-layer HBM3E volume production

SK hynix began volume production of 36 GB, 12-layer HBM3E.

Source

Meta Llama 3.2 vision and edge models

Meta released Llama 3.2 with vision models and small models for phones.

Source

CSCS Alps supercomputer inaugurated

ETH Zurich inaugurated the Alps supercomputer at the Swiss National Supercomputing Centre in Lugano.

Source

OpenAI OpenAI o1-preview and o1-mini

OpenAI released o1-preview and o1-mini, reasoning models that think before answering.

Source

Netherlands Netherlands widens licensing to more ASML tools

The Dutch government announced, effective the next day, an expanded national export control on advanced chipmaking equipment.

Source

xAI xAI brings Colossus online

Elon Musk said xAI's Colossus cluster of 100,000 Nvidia GPUs had come online.

Source

Cerebras Cerebras Inference

Cerebras launched an inference service, claiming 1,800 tokens per second on Llama 3.1 8B.

Source

OpenAI Structured Outputs in the API

OpenAI added Structured Outputs so model responses reliably match a developer-supplied JSON schema.

Source

Midjourney Midjourney V6.1

Midjourney released V6.1, available on Discord and on its website.

Source

OpenAI SearchGPT prototype

OpenAI tested SearchGPT, a search prototype with sourced answers, with a small group of users.

Source

Meta Llama 3.1 405B

Meta released Llama 3.1, including a 405-billion-parameter open-weight model.

Source

Tenstorrent Tenstorrent puts Wormhole cards on sale

Tenstorrent opened orders for Wormhole n150s and n300s PCIe cards and workstations aimed at developers.

Source

OpenAI GPT-4o mini

OpenAI released GPT-4o mini, a small low-cost model that replaced GPT-3.5 Turbo in ChatGPT.

Source

SoftBank SoftBank buys Graphcore

SoftBank acquired British AI chip designer Graphcore, with the price undisclosed.

Source

Etched Etched announces Sohu, a transformer-only chip

The startup Etched said it raised $120 million for Sohu, a chip built only for transformer models.

Source

Anthropic Claude 3.5 Sonnet and Artifacts

Anthropic released Claude 3.5 Sonnet and Artifacts, a side panel for generated code and documents.

Source

Apple Apple Intelligence announced

At WWDC Apple announced Apple Intelligence, on-device and private-cloud AI features, plus optional ChatGPT integration.

Source

Nvidia Nvidia names Rubin as next platform

Ahead of Computex, Nvidia revealed for the first time that the Rubin platform, with a new Arm-based CPU called Vera, would succeed Blackwell.

Source

US Department of Energy Aurora passes one exaflop

Argonne's Aurora, built on Intel processors and GPUs, became the second system to break the exascale barrier on the TOP500, at 1.012 exaflops.

Source

UALink Consortium UALink promoter group formed

Industry companies formed the Ultra Accelerator Link promoter group to develop an open link between AI accelerators.

Source

Microsoft Copilot+ PCs

Microsoft announced Copilot+ PCs with on-device AI hardware.

Source

Google Google Trillium, sixth-generation TPU

Google announced Trillium, its sixth TPU generation, claiming 4.7 times the peak compute per chip of TPU v5e.

Source

Google Google I/O 2024: Veo, Imagen 3, AI Overviews

Google announced the Veo video model, Imagen 3, Gemini 1.5 Flash and AI Overviews in US Search.

Source

OpenAI GPT-4o released

OpenAI introduced GPT-4o, a natively multimodal model handling text, audio and images, and made it available to free users.

Source

OpenAI Model Spec first published

OpenAI published a draft Model Spec describing how its models should behave and resolve conflicts.

Source

Apple Apple M4

Apple introduced the M4, with a Neural Engine rated at up to 38 trillion operations per second.

Source

Meta Llama 3 and Meta AI assistant

Meta released Llama 3 in 8B and 70B sizes and upgraded its Meta AI assistant.

Source

Meta Meta details its next-generation MTIA chip

Meta described a TSMC 5 nm successor to MTIA, already in 16 regions, with a 90 W design and up to three times the performance of the first chip on four models.

Source

Intel Intel Gaudi 3

Intel introduced Gaudi 3 with 128 GB of HBM memory and, by its own projections, about 50 percent faster training than the H100 on selected models.

Source

Google TPU v5p reaches general availability

Google said at Next '24 that TPU v5p was generally available.

Source

OpenAI ChatGPT without an account

OpenAI began letting people use ChatGPT without signing up.

Source

Nvidia Nvidia Blackwell, B200 and GB200

Nvidia announced the Blackwell platform, with 208 billion transistors per GPU on TSMC's 4NP process, along with the B200 GPU and the GB200 Grace Blackwell Superchip.

Source

Nvidia Nvidia announces X800 800 Gb/s switches

Nvidia announced Quantum-X800 InfiniBand and Spectrum-X800 Ethernet, which it called the first end-to-end 800 Gb/s networking platforms.

Source

xAI Grok-1 weights released

xAI released the 314B-parameter Grok-1 base model weights under Apache 2.0.

Source

Cerebras Cerebras WSE-3

Cerebras announced the 5-nanometre WSE-3, with 4 trillion transistors and 900,000 cores.

Source

Meta Meta details 24,576-GPU clusters

Meta shared details of two 24,576-GPU clusters supporting Llama 3 development, one on RoCE Ethernet and one on InfiniBand.

Source

JEDEC JEDEC publishes GDDR7

JEDEC published JESD239, the GDDR7 standard, which doubles GDDR6 bandwidth to as much as 192 GB/s per device.

Source

Anthropic Claude 3 family

Anthropic released Claude 3 Opus, Sonnet and Haiku, adding image input.

Source

Samsung Samsung 36 GB HBM3E 12H

Samsung announced what it called the industry's first 36 GB, 12-layer HBM3E.

Source

Micron Micron starts HBM3E volume production

Micron said it had begun volume production of HBM3E memory.

Source

Mistral Mistral Large and Le Chat

Mistral launched Mistral Large and its Le Chat assistant, plus a Microsoft partnership.

Source

Google Gemma open models

Google released Gemma, lightweight open-weight models built from Gemini research.

Source

DIPC DIPC's Hyperion supercomputer shown in Spain

A Basque minister visited Hyperion at the Donostia physics centre, a 2.8-million-euro system with over 14,000 cores.

Source

Nvidia Nvidia reveals Eos

Nvidia gave a look at Eos, its in-house supercomputer, ranked in the world's top 10.

Source

Google Gemini 1.5 Pro

Google announced Gemini 1.5 Pro with a context window of up to one million tokens.

Source

OpenAI Sora text-to-video preview

OpenAI previewed Sora, a model that makes up to minute-long videos from text, to red teamers and selected artists.

Source

OpenAI ChatGPT memory testing begins

OpenAI started testing memory, letting ChatGPT remember details across conversations, with user controls.

Source

Google Bard renamed Gemini; Gemini Advanced

Google rebranded Bard as Gemini and launched Gemini Advanced with Ultra 1.0 in a paid plan.

Source

OpenAI GPT Store and ChatGPT Team

OpenAI opened the GPT Store for sharing custom GPTs and launched the ChatGPT Team plan for small groups.

Source

Midjourney Midjourney V6

Midjourney released V6 in alpha with better prompt following and text.

Source

Intel Intel launches Core Ultra with an NPU

Intel launched Core Ultra, its first consumer chips with a neural processing unit, at its AI Everywhere event.

Source

Mistral Mixtral 8x7B

Mistral released Mixtral 8x7B, an open sparse mixture-of-experts model, after posting the weights by torrent.

Source

AMD AMD Instinct MI300X and MI300A

AMD launched the MI300 series, including the MI300X with 192 GB of HBM3 memory and the MI300A, which combines CPU and GPU.

Source

Google Google announces Cloud TPU v5p and AI Hypercomputer

Google announced TPU v5p, with pods of 8,960 chips, which it said train large language models 2.8 times faster than TPU v4.

Source

Google Gemini 1.0 announced

Google DeepMind introduced Gemini in Ultra, Pro and Nano sizes; Pro began powering Bard the same day.

Source

JEDEC JEDEC publishes the CAMM2 module standard

JEDEC published JESD318, a compression-attached module standard covering DDR5 and LPDDR5/5X.

Source

Amazon Web Services AWS announces Trainium2

AWS announced Trainium2, its second training chip, at re:Invent.

Source

OpenAI Sam Altman agreed to return as CEO

OpenAI reached an agreement for Altman to return as CEO with a new initial board.

Source

Anthropic Claude 2.1

Anthropic released Claude 2.1 with a 200,000-token context window and beta tool use.

Source

OpenAI Board removes Sam Altman as CEO

OpenAI's board announced Sam Altman's departure as CEO; he returned days later after staff and investor pressure.

Source

Microsoft Microsoft Maia 100

Microsoft introduced Maia 100, its first custom AI accelerator, at Ignite.

Source

Nvidia Nvidia H200

Nvidia announced the H200, which it described as the first GPU with HBM3e memory.

Source

OpenAI GPTs: custom versions of ChatGPT

OpenAI let users build custom ChatGPT versions with instructions, knowledge files and actions, without code.

Source

OpenAI GPT-4 Turbo preview

OpenAI previewed GPT-4 Turbo with a 128,000-token context window and cheaper pricing.

Source

OpenAI OpenAI DevDay 2023

At its first developer conference OpenAI announced GPT-4 Turbo with 128K context, the Assistants API, GPTs and lower prices.

Source

OpenAI Assistants API beta

OpenAI released a beta API for building assistants with threads, retrieval and code interpreter tools.

Source

xAI Grok announced

xAI unveiled Grok, a chatbot for X Premium+ subscribers.

Source

Apple Apple M3 family

Apple unveiled the M3, M3 Pro and M3 Max, with up to 128 GB of unified memory.

Source

US Department of Commerce US tightens chip export controls, October 2023

The Federal Register published US Commerce rules tightening controls on advanced computing chips and chipmaking equipment.

Source

Mistral Mistral 7B

Mistral AI released its first model, Mistral 7B, under Apache 2.0.

Source

Anthropic Amazon to invest up to $4 billion in Anthropic

Amazon agreed to invest up to $4 billion in Anthropic for a minority stake.

Source

OpenAI ChatGPT can see, hear and speak

OpenAI announced voice conversations and image input (GPT-4V) for ChatGPT Plus and Enterprise.

Source

European Union European Chips Act enters into force

The European Chips Act took effect, with a goal of doubling the EU's global chip market share to 20% by 2030.

Source

OpenAI DALL-E 3 announced

OpenAI announced DALL-E 3, built to follow prompts closely and to be used through ChatGPT.

Source

Netherlands Dutch chip-equipment licence rule takes effect

The Dutch national authorisation requirement for certain advanced chipmaking equipment entered into force.

Source

Google Google Cloud TPU v5e

Google announced Cloud TPU v5e, a chip aimed at cost-efficient training and inference.

Source

OpenAI ChatGPT Enterprise

OpenAI launched a business plan with enterprise security, unlimited GPT-4 and admin tools.

Source

Meta Code Llama

Meta released Code Llama, Llama 2 models specialized for programming.

Source

OpenAI Fine-tuning for GPT-3.5 Turbo

Developers could fine-tune GPT-3.5 Turbo on their own data.

Source

SK hynix SK hynix HBM3E samples

SK hynix said it had developed HBM3E and was providing samples to a customer for evaluation.

Source

Alibaba (Qwen) Qwen-7B open weights

Alibaba released Qwen-7B as open weights.

Source

OpenAI Custom instructions in ChatGPT

ChatGPT gained custom instructions so users can set preferences that apply to every conversation.

Source

Ultra Ethernet Consortium Ultra Ethernet Consortium founded

AMD, Arista, Broadcom, Cisco, Eviden, HPE, Intel, Meta and Microsoft formed a Linux Foundation group to adapt Ethernet to AI and high-performance computing.

Source

Meta Llama 2

Meta released Llama 2 with a licence allowing commercial use, in partnership with Microsoft.

Source

Anthropic Claude 2 and claude.ai

Anthropic released Claude 2 and opened the claude.ai chat site to the public in the US and UK.

Source

Open source Ollama first release

Ollama released a tool for downloading and running open models locally.

Source

OpenAI GPT-4 API generally available; Code Interpreter rolls out

OpenAI opened GPT-4 to all paying API users and began rolling out Code Interpreter to ChatGPT Plus.

Source

China China announces gallium and germanium export controls

China's commerce ministry and customs announced export licences for gallium, germanium and related compounds, from August 1.

Source

Netherlands Dutch controls on chipmaking equipment

The Dutch government published export controls requiring a licence for advanced chipmaking equipment, effective September 1, 2023.

Source

OpenAI Function calling in the API

OpenAI added function calling and new GPT-4 and GPT-3.5-turbo versions with lower prices.

Source

Apple Apple M2 Ultra

Apple introduced the M2 Ultra, with up to 192 GB of unified memory and 800 GB per second of bandwidth.

Source

Nvidia Nvidia Spectrum-X Ethernet for AI

Nvidia launched Spectrum-X, an Ethernet networking platform designed for generative AI data centres.

Source

Nvidia Grace Hopper in production and the DGX GH200

Nvidia announced the DGX GH200 AI supercomputer, powered by its GH200 Grace Hopper Superchips.

Source

Meta Meta MTIA v1

Meta described MTIA v1, an in-house inference accelerator made on TSMC's 7-nanometre process.

Source

OpenAI ChatGPT iOS app

OpenAI released an official ChatGPT app for iPhone in the US.

Source

OpenAI Web browsing and plugins beta for Plus

OpenAI began rolling out web browsing and plugins to all ChatGPT Plus subscribers as beta features.

Source

Anthropic 100K context window

Anthropic expanded Claude's context window to 100,000 tokens.

Source

Google PaLM 2 at Google I/O

Google announced PaLM 2, powering Bard and many Google products, and opened Bard without a waitlist.

Source

OpenAI Option to turn off chat history

ChatGPT users gained a setting to stop conversations being used for training and kept in history.

Source

Amazon Web Services AWS Inf2 instances generally available

AWS made EC2 Inf2 instances, built on Inferentia2, generally available.

Source

Alibaba (Qwen) Alibaba unveils Tongyi Qianwen

Alibaba Cloud announced its Tongyi Qianwen (Qwen) chatbot model for business testing.

Source

Google Google details TPU v4 optical switching

Google published details on TPU v4 pods, joined by its own optical circuit switches, and said the chip beat TPU v3 by 2.1 times per chip.

Source

Open source AutoGPT released

Toran Bruce Richards released AutoGPT, an open-source agent that chains GPT-4 calls toward goals.

Source

OpenAI ChatGPT plugins announced

OpenAI introduced plugins, including web browsing and a code interpreter, to a limited set of users.

Source

Google Bard opens to US and UK users

Google began giving users in the US and UK access to Bard through a waitlist.

Source

Microsoft Microsoft 365 Copilot announced

Microsoft announced Copilot for Word, Excel, PowerPoint, Outlook and Teams.

Source

Midjourney Midjourney V5

Midjourney released V5 in alpha with more photorealistic output.

Source

OpenAI GPT-4 Technical Report on arXiv

OpenAI posted the GPT-4 technical report, giving benchmark results but few architecture details.

Source

Anthropic Claude launched

Anthropic released Claude and Claude Instant to businesses through its API and partners.

Source

OpenAI GPT-4 released

OpenAI announced GPT-4, a large multimodal model, available in ChatGPT Plus and via an API waitlist.

Source

Open source llama.cpp released

Georgi Gerganov released llama.cpp, letting LLaMA models run on ordinary laptops in C/C++.

Source

OpenAI ChatGPT (gpt-3.5-turbo) and Whisper APIs

OpenAI opened the gpt-3.5-turbo chat model and Whisper to developers at much lower prices than earlier models.

Source

Meta LLaMA released to researchers

Meta released LLaMA, a family of 7B to 65B models, under a research-only licence; the weights soon leaked.

Source

Microsoft New Bing with chat

Microsoft unveiled an AI-powered Bing and Edge with a chat assistant built on an OpenAI model.

Source

Google Bard announced

Google announced Bard, an experimental conversational AI service based on LaMDA, for trusted testers.

Source

Quora (Poe) Poe opens to the public

Quora opened Poe, an app for chatting with several AI bots, to everyone on iOS.

Source

OpenAI ChatGPT Plus subscription

OpenAI introduced a $20 monthly plan with priority access and faster responses.

Source

OpenAI Microsoft extends OpenAI partnership

Microsoft announced a multiyear, multibillion-dollar extension of its investment in OpenAI.

Source

AMD AMD brings Ryzen AI to laptop chips

At CES 2023, AMD announced Ryzen 7000 mobile processors that bring Ryzen AI technology to select laptops.

Source

TSMC TSMC starts 3 nm volume production

TSMC held a ceremony marking 3-nanometre volume production at Fab 18 in southern Taiwan.

Source

Anthropic Constitutional AI paper

Anthropic described training a harmless assistant using a written set of principles and AI feedback.

Source

OpenAI ChatGPT launched as a research preview

OpenAI released ChatGPT, a free conversational interface built on a GPT-3.5 model fine-tuned for dialogue.

Source

Midjourney Midjourney V4 alpha

Midjourney released its V4 model in alpha.

Source

US Bureau of Industry and Security US publishes first export limits on advanced AI chips

BIS published rules requiring licences for advanced computing chips above set speed and data-rate limits bound for China.

Source

Amazon Web Services AWS Trn1 instances generally available

AWS made EC2 Trn1 instances, built on Trainium, generally available.

Source

US Department of Commerce US export controls on advanced chips, October 2022

The US Commerce Department issued rules restricting exports of advanced computing chips and chipmaking equipment to China.

Source

OpenAI DALL-E 2 waitlist removed

OpenAI opened DALL-E 2 to everyone without a waitlist.

Source

OpenAI Whisper speech recognition open-sourced

OpenAI released Whisper, an open-source multilingual speech recognition model.

Source

Nvidia GeForce RTX 40 series (Ada Lovelace)

Nvidia announced the GeForce RTX 40 series, led by the RTX 4090, on its Ada Lovelace architecture.

Source

Nvidia Nvidia H100 enters full production

Nvidia said the H100 was in full production, with partners planning first products and services in October.

Source

Character.AI Character.AI public beta

Character.AI opened a public beta for chatting with user-created characters.

Source

Nvidia US imposes licence on Nvidia A100 and H100

Nvidia disclosed that the US government had imposed a licence requirement on exports of A100 and H100 chips to China and Russia.

Source

Stability AI Stable Diffusion public release

Stability AI and partners released Stable Diffusion with open weights.

Source

Broadcom Broadcom ships Tomahawk 5 at 51.2 Tb/s

Broadcom delivered a single Ethernet switch chip with 51.2 Tb/s of capacity, double any other on the market, by its account.

Source

US government CHIPS and Science Act signed

The United States enacted the CHIPS and Science Act, funding domestic semiconductor manufacturing and research.

Source

CXL Consortium CXL 3.0 specification released

The CXL Consortium released version 3.0, adding fabric capabilities and better memory sharing and pooling.

Source

OpenAI DALL-E 2 enters beta

OpenAI began inviting a million waitlisted users to DALL-E 2 with a credit-based pricing model.

Source

Midjourney Midjourney open beta

Midjourney opened its Discord-based image generator to the public.

Source

EuroHPC JU LUMI inaugurated in Finland

EuroHPC inaugurated LUMI in Kajaani, ranked third on the TOP500.

Source

US Department of Energy Frontier breaks the exascale barrier

Oak Ridge's Frontier, built on AMD EPYC processors and Instinct accelerators, became the first system on the TOP500 to pass one exaflop, at 1.1 exaflops.

Source

Google Imagen text-to-image research

Google Research presented Imagen, a text-to-image diffusion model, without public release.

Source

Intel Intel's Habana launches Gaudi2

Habana Labs launched the 7 nm Gaudi2 training chip at Intel Vision, with 96 GB of on-board memory and 24 Ethernet ports.

Source

OpenAI DALL-E 2 announced

OpenAI introduced DALL-E 2, producing more realistic, higher-resolution images and image edits from text.

Source

Google PaLM announced

Google described PaLM, a 540-billion-parameter language model trained with its Pathways system.

Source

Nvidia Nvidia Hopper and the H100

Nvidia announced the Hopper architecture and the H100 GPU, with 80 billion transistors on TSMC's 4N process.

Source

Apple Apple M1 Ultra

Apple unveiled the M1 Ultra, with up to 128 GB of unified memory.

Source

AMD AMD completes Xilinx purchase

AMD completed its all-stock acquisition of Xilinx, adding adaptive chips to its data centre portfolio.

Source

Nvidia Nvidia and SoftBank drop the Arm deal

Nvidia and SoftBank ended the Arm purchase, citing regulatory obstacles, and SoftBank said it would explore a public offering of Arm.

Source

JEDEC JEDEC publishes HBM3 standard

JEDEC published the HBM3 standard, defining data rates up to 6.4 gigabits per second per pin, or 819 GB per second per device.

Source

OpenAI InstructGPT becomes API default

OpenAI made instruction-following models trained with human feedback (RLHF) the default in its API.

Source

Meta Meta AI Research SuperCluster

Meta introduced its AI Research SuperCluster, an AI supercomputer.

Source

OpenAI OpenAI API waitlist removed

OpenAI made its API available to developers in supported countries without a waitlist.

Source

AMD AMD Instinct MI200 series (MI250X)

AMD announced the Instinct MI200 series, including the MI250X, for exascale systems.

Source

SK hynix SK hynix develops HBM3

SK hynix announced it had developed HBM3 memory, in 16 GB and 24 GB versions.

Source

Tesla Tesla reveals the D1 chip for Dojo

At AI Day, Tesla showed its D1 training chip, a 7 nm design rated at 362 teraflops, with 25 chips forming one training tile.

Source

OpenAI OpenAI Codex in the API

OpenAI released an improved Codex model to API beta users, able to turn natural language into code.

Source

OpenAI Codex paper: Evaluating LLMs Trained on Code

OpenAI published research on Codex, a GPT model fine-tuned on public code, and introduced the HumanEval benchmark.

Source

US Department of Energy Berkeley Lab dedicates Perlmutter

NERSC dedicated Perlmutter, an HPE Cray system whose first phase has 1,536 nodes, each with four Nvidia A100 GPUs.

Source

Google Google TPU v4

At Google I/O, Sundar Pichai announced TPU v4, saying one pod delivers more than one exaflop.

Source

Google LaMDA unveiled

Google showed LaMDA, a conversational language model, at Google I/O.

Source

Nvidia Nvidia announces the Grace CPU

Nvidia announced Grace, its first data centre CPU, for giant AI and high-performance computing workloads.

Source

Cerebras Cerebras WSE-2

Cerebras unveiled the WSE-2, with 2.6 trillion transistors and 850,000 cores.

Source

OpenAI DALL-E announced

OpenAI showed a GPT-3-style model that creates images from text descriptions.

Source

OpenAI CLIP released

OpenAI released CLIP, a model that learns to link images and text from internet data.

Source

US Department of Commerce US adds SMIC to the Entity List

The US added China's top chipmaker, SMIC, and other entities to the Entity List, citing China's military-civil fusion doctrine.

Source

Amazon Web Services AWS announces Trainium

At re:Invent, AWS announced Trainium, its custom chip for machine learning training.

Source

AMD AMD Instinct MI100

AMD announced the Instinct MI100 accelerator on its CDNA architecture, claiming more than 10 teraflops of double-precision performance.

Source

Nvidia Nvidia introduces 400G InfiniBand

Nvidia introduced its next generation of Mellanox InfiniBand at 400 Gb/s at SC20.

Source

Apple Apple M1 and unified memory

Apple unveiled the M1, its first Mac chip, with a unified memory architecture and a 16-core Neural Engine rated at 11 trillion operations per second.

Source

Nvidia Nvidia launches the BlueField-2 DPU family

Nvidia announced data processing units and the DOCA software kit, moving networking, storage and security work off the CPU.

Source

Nvidia Nvidia agrees to buy Arm for $40 billion

Nvidia announced a deal to buy the chip designer Arm from SoftBank for $40 billion.

Source

Nvidia GeForce RTX 30 series

Nvidia announced the GeForce RTX 3090, 3080 and 3070, built on Ampere.

Source

JEDEC JEDEC publishes DDR5

JEDEC published JESD79-5, the DDR5 standard, promising twice the performance of the previous generation for servers and client systems.

Source

SK hynix SK hynix starts HBM2E mass production

SK hynix began mass production of HBM2E, at 16 GB per stack and over 460 GB/s.

Source

RIKEN Fugaku takes first place on the TOP500

RIKEN said Fugaku, built with Fujitsu on Arm technology, had taken first place on the TOP500 with 415.53 petaflops.

Source

OpenAI OpenAI API launched (GPT-3 private beta)

OpenAI opened a commercial API giving waitlisted developers access to GPT-3 models.

Source

Nvidia Nvidia's Selene debuts at seventh on the TOP500

Selene, built from DGX A100 systems with A100 GPUs, entered the June 2020 list at seventh with 27.58 petaflops.

Source

OpenAI GPT-3 paper: Language Models are Few-Shot Learners

OpenAI described the 175-billion-parameter GPT-3 and showed it could do tasks from a few examples in the prompt.

Source

Microsoft Microsoft's supercomputer for OpenAI

Microsoft announced a supercomputer built with OpenAI, with more than 285,000 CPU cores and 10,000 GPUs.

Source

US Department of Commerce US foreign direct product rule on Huawei

The US amended its export rules so that some chips made abroad with US technology for Huawei designs need a licence.

Source

TSMC TSMC announces Arizona fab

TSMC said it intended to build a 5-nanometre chip plant in Arizona.

Source

Nvidia Nvidia Ampere and the A100

Nvidia announced the Ampere architecture and the A100 data centre GPU, with more than 54 billion transistors on a 7-nanometre process, in full production.

Source

OpenAI Jukebox music generation model

OpenAI released a model that generates raw audio songs, including rudimentary singing.

Source

Nvidia Nvidia completes Mellanox purchase

Nvidia completed its acquisition of Mellanox.

Source

TSMC TSMC starts 5 nm volume production

TSMC says it was the first foundry to move 5-nanometre FinFET into volume production, in 2020.

Source

Intel Intel buys Habana Labs

Intel bought Israeli AI chipmaker Habana Labs for about 2 billion US dollars.

Source

Amazon Web Services AWS Inf1 instances

AWS launched EC2 Inf1 instances, the first cloud servers using its Inferentia chips.

Source

OpenAI Full GPT-2 (1.5B) released

OpenAI completed the staged release by publishing the full 1.5-billion-parameter GPT-2 weights.

Source

Google BERT used in Google Search

Google said it was using BERT to better understand search queries.

Source

Alibaba (Qwen) Alibaba unveils the Hanguang 800

Alibaba's chip unit Pingtouge unveiled its first AI inference chip at the Apsara Conference, for use in Alibaba's cloud.

Source

Huawei Huawei Ascend 910

Huawei launched the Ascend 910 AI processor, rated at 256 teraflops, and the MindSpore framework.

Source

OpenAI GPT-2 774M model released

As part of the staged release, OpenAI published the 774-million-parameter GPT-2 model with a six-month report.

Source

Cerebras Cerebras Wafer Scale Engine

Cerebras unveiled the Wafer Scale Engine, a single chip with more than 1.2 trillion transistors and an area of 46,225 square millimetres.

Source

OpenAI Microsoft invests $1 billion in OpenAI

Microsoft agreed to invest one billion dollars and become OpenAI's exclusive cloud provider.

Source

Japan Japan tightens chip-material exports to South Korea

Japan's trade ministry moved fluorinated polyimide, resists and hydrogen fluoride bound for South Korea to individual export licences.

Source

Nvidia Nvidia DGX SuperPOD reaches 22nd on the TOP500

Nvidia described a cluster of 96 DGX-2H servers and 1,536 V100 GPUs that scored 9.444 petaflops and placed 22nd on the June 2019 list.

Source

US Department of Commerce US adds Huawei to the Entity List

The US Commerce Department added Huawei and affiliates to the Entity List, requiring licences for US technology exports to them.

Source

OpenAI MuseNet music model

OpenAI showed a transformer that composes multi-instrument music in several styles.

Source

Nvidia Nvidia announces the $99 Jetson Nano

Nvidia announced a small $99 AI computer that delivers 472 GFLOPS.

Source

Nvidia Nvidia agrees to buy Mellanox

Nvidia agreed to acquire Mellanox, a maker of InfiniBand and Ethernet data centre networking, for 6.9 billion US dollars.

Source

OpenAI OpenAI LP capped-profit structure

OpenAI created a capped-profit company under its non-profit to raise outside capital.

Source

Google Google opens Coral, with the Edge TPU, in public beta

Google put Coral in public beta, offering local AI hardware built on its small Edge TPU chip.

Source

JEDEC JEDEC publishes LPDDR5

JEDEC published JESD209-5, a low-power memory standard targeting 6,400 MT/s.

Source

OpenAI GPT-2 announced with staged release

OpenAI unveiled the 1.5-billion-parameter GPT-2 but released only a small version, citing misuse concerns.

Source

Amazon Web Services AWS announces Inferentia

AWS announced Inferentia, a custom machine learning inference chip.

Source

AMD AMD Radeon Instinct MI60 and MI50

AMD unveiled the MI60 and MI50, which it called the first 7-nanometre data centre GPUs.

Source

Google BERT paper

Google introduced BERT, a bidirectional transformer pre-trained on text, which set records on language understanding tasks.

Source

Nvidia GeForce RTX 20 series (Turing)

Nvidia announced the GeForce RTX 2080 Ti, 2080 and 2070, its first consumer RTX cards.

Source

OpenAI GPT-1: generative pre-training paper

OpenAI described a transformer language model pre-trained on unlabeled text and then fine-tuned for specific tasks, the first model later called GPT.

Source

US Department of Energy Summit supercomputer launched

Oak Ridge launched Summit, with 27,648 Nvidia V100 GPUs and a peak of 200 petaflops.

Source

Nvidia Nvidia HGX-2 platform

Nvidia introduced HGX-2, a platform for up to 16 GPUs; the DGX-2 was the first system built on it.

Source

Google Google TPU 3.0

At Google I/O, Sundar Pichai announced the third-generation TPU, with pods he said were eight times more powerful, and liquid cooling in Google's data centres.

Source

Nvidia Nvidia DGX-2 and NVSwitch

At GTC, Nvidia unveiled the DGX-2, a system joining 16 GPUs through its new NVSwitch fabric, which it rated at 2 petaflops.

Source

Google Cloud TPUs open in beta

Google made Cloud TPUs available in beta on Google Cloud, so outside customers could rent its accelerators.

Source

TSMC TSMC starts 7 nm volume production

TSMC says it was the first foundry to start 7-nanometre FinFET volume production, in 2018.

Source

Nvidia Nvidia Titan V

Nvidia introduced the Titan V, a Volta-based GPU for desktop PCs aimed at AI researchers.

Source

Apple Apple A11 Bionic with Neural Engine

Apple announced the iPhone X with the A11 Bionic chip and its Neural Engine.

Source

Google Attention Is All You Need

Google researchers introduced the Transformer architecture, which underlies modern large language models.

Source

Google Second-generation Cloud TPU

Google announced second-generation Cloud TPUs, each rated at up to 180 teraflops, for both training and inference.

Source

Nvidia Nvidia launches Volta and the Tesla V100

Nvidia launched the Volta architecture and the Tesla V100 data centre GPU, which it rated above 120 teraflops of deep learning performance.

Source

Google Google publishes TPU performance paper

Google researchers published a paper analysing the performance of the first TPU in its data centres.

Source

AMD AMD Radeon Instinct MI6, MI8 and MI25

AMD introduced Radeon Instinct, its accelerator brand for machine intelligence, with the MI6, MI8 and MI25 and the MIOpen library.

Source

AMD AMD releases ROCm 1.3 at SC16

AMD released a new version of its open-source Radeon Open Compute platform, adding Polaris GPU support, Docker containers and a CUDA porting path through HIP.

Source

Research Google Neural Machine Translation

Google announced a neural machine translation system, first used for Chinese to English in Google Translate, replacing its phrase-based approach.

Source

Paper DeepMind's WaveNet

DeepMind introduced WaveNet, a neural network that generates raw audio waveforms and produced more natural synthetic speech.

Source

Intel Intel buys Nervana Systems

Intel announced it was buying the deep learning startup Nervana, which was designing a custom training chip; press reports valued the deal above $350 million.

Source

China Sunway TaihuLight tops the TOP500

China's Sunway TaihuLight took first place at 93 petaflops, using processors designed and made in China.

Source

Google Google reveals the TPU

Google revealed the Tensor Processing Unit, a custom chip for machine learning that it said was already running in its data centres.

Source

Nvidia Nvidia DGX-1 and Tesla P100 (Pascal)

Nvidia launched the DGX-1 deep learning system, built on Pascal-based Tesla P100 GPUs.

Source

Program Microsoft's Tay chatbot

Microsoft launched Tay, a Twitter chatbot that learned from users; it was taken offline within a day after being manipulated into posting offensive messages.

Source

Event AlphaGo defeats Lee Sedol

DeepMind's AlphaGo beat Go champion Lee Sedol four games to one in Seoul, a milestone many experts had expected to be a decade away.

Source

Paper AlphaGo paper and win over Fan Hui

DeepMind published AlphaGo in Nature, revealing it had beaten European champion Fan Hui 5–0 in October 2015, the first win by a program over a professional Go player on a full board without handicap.

Source

Samsung Samsung starts mass-producing HBM2

Samsung said it had begun mass-producing the industry's first 4 GB DRAM package using the HBM2 interface.

Source

JEDEC JEDEC updates the HBM standard (HBM2)

JEDEC published JESD235A, an update to the High Bandwidth Memory standard, supporting up to 8 GB per device at up to 256 GB per second.

Source

OpenAI OpenAI founded

OpenAI was announced as a non-profit AI research lab with backing from several tech investors.

Source

Paper Residual networks (ResNet)

Kaiming He and colleagues at Microsoft Research introduced residual connections that made very deep networks trainable, winning ImageNet 2015.

Source

AMD AMD launches the Boltzmann Initiative

AMD announced a compiler, a headless Linux driver and the HIP tool for porting CUDA code, aimed at GPU computing on its FirePro cards.

Source

Invention Google open-sources TensorFlow

Google released TensorFlow, its machine learning software library, as open source.

Source

AMD AMD Radeon Fury, first GPU family with HBM

AMD announced its Radeon 300 series, including what it called the world's first graphics family with High Bandwidth Memory.

Source

Program Amazon Echo and Alexa

Amazon introduced the Echo speaker with its voice assistant Alexa, initially by invitation.

Source

Paper Sequence-to-sequence learning

Ilya Sutskever, Oriol Vinyals and Quoc Le showed that LSTM networks could map one sequence to another, achieving strong machine translation results.

Source

Paper Attention mechanism for neural translation

Dzmitry Bahdanau, Kyunghyun Cho and Yoshua Bengio added an attention mechanism letting a translation network focus on relevant source words, a precursor to transformers.

Source

Paper Generative adversarial networks

Ian Goodfellow and colleagues introduced GANs, in which two networks compete, one generating samples and one judging them, to produce realistic data.

Source

Event Eugene Goostman Turing test claim

Organisers at the University of Reading claimed the chatbot Eugene Goostman passed a Turing test by convincing a third of judges; many researchers disputed the claim.

Source

Nvidia Nvidia announces NVLink

Nvidia announced NVLink, a high-speed GPU interconnect co-developed with IBM, and said it would arrive in the Pascal architecture expected in 2016.

Source

Event Google acquires DeepMind

Google agreed to buy DeepMind, reportedly for around 400 to 650 million U.S. dollars.

Source

Paper DeepMind's deep Q-network plays Atari

DeepMind showed a single neural network learning to play several Atari games directly from screen pixels using reinforcement learning.

Source

JEDEC JEDEC publishes the first HBM standard

JEDEC published JESD235, the first standard for high-bandwidth memory, DRAM stacked close to the processor over a wide interface.

Source

Paper word2vec

Tomas Mikolov and colleagues at Google published word2vec, an efficient way to represent words as vectors that capture meaning from context.

Source

Nvidia Nvidia Tesla K20 and the Titan supercomputer

Nvidia announced the Kepler-based Tesla K20 accelerators, which it said power Titan at Oak Ridge, ranked first in the world.

Source

US Department of Energy Oak Ridge debuts Titan

Oak Ridge debuted Titan, a Cray system with 18,688 nodes, each with an AMD Opteron and an Nvidia Tesla K20 GPU.

Source

Research Google Brain learns to recognise cats

A Google and Stanford team trained a large neural network on unlabelled YouTube frames; it learned features responding to faces and cats.

Source

Research AlexNet wins ImageNet

Alex Krizhevsky, Ilya Sutskever and Geoffrey Hinton's convolutional network, trained on GPUs, won the ImageNet challenge by a wide margin, starting the deep learning boom.

Source

Program Siri launches on the iPhone 4S

Apple announced Siri, a voice assistant built on technology from a startup it had acquired, as a feature of the iPhone 4S.

Source

Event IBM Watson wins Jeopardy!

IBM's Watson question-answering system beat champions Ken Jennings and Brad Rutter on the television quiz show Jeopardy!.

Source

China China's Tianhe-1A reaches 2.5 petaflops with GPUs

China unveiled Tianhe-1A in Tianjin, with 7,168 Nvidia Tesla M2050 GPUs and a sustained 2.5 petaflops.

Source

Event DeepMind founded

Demis Hassabis, Shane Legg and Mustafa Suleyman founded DeepMind in London to research general-purpose learning systems.

Source

Nvidia Nvidia unveils the Fermi architecture

Nvidia introduced its next CUDA architecture, codenamed Fermi, built for GPU computing as well as graphics.

Source

Research ImageNet dataset

Fei-Fei Li and colleagues presented ImageNet, a database of millions of labelled images that became the main benchmark for computer vision.

Source

Lawrence Livermore National Laboratory Hyperion testbed announced at Lawrence Livermore

Lawrence Livermore and 10 computing companies announced Hyperion, a Linux cluster testbed for next-generation supercomputing.

Source

US Department of Energy Roadrunner passes one petaflop

Los Alamos's IBM Roadrunner became the first system to pass one petaflop, reaching 1.026 and taking first place on the TOP500.

Source

Nvidia Nvidia launches the Tesla line

Nvidia launched Tesla, a line of GPU computing processors aimed at scientific and technical computing rather than graphics.

Source

Nvidia Nvidia unveils CUDA

Nvidia unveiled CUDA, a computing architecture for its GPUs with what it called the industry's first C compiler environment for the GPU.

Source

Paper Hinton's deep belief networks

Geoffrey Hinton, Simon Osindero and Yee-Whye Teh showed how to train deep neural networks one layer at a time, a turning point for what became known as deep learning.

Source

Event Stanford's Stanley wins the DARPA Grand Challenge

Five autonomous vehicles completed the 212 km desert course, with Stanford's Stanley finishing first.

Source

Event First DARPA Grand Challenge

DARPA held a race for driverless vehicles in the Mojave Desert; no vehicle finished, and the best travelled about 12 km.

Source

Invention iRobot's Roomba

iRobot launched the Roomba vacuum cleaner, which became the first widely sold home robot.

Source

Invention The Wayback Machine opens to the public

The Internet Archive unveiled the Wayback Machine at UC Berkeley's Bancroft Library, letting anyone view more than 10 billion archived web pages saved since 1996.

Source

InfiniBand Trade Association InfiniBand 1.0 specification

The InfiniBand Trade Association, founded in 1999, released version 1.0 of the InfiniBand Architecture specification.

Source

Invention Honda's ASIMO

Honda unveiled ASIMO, a humanoid robot able to walk smoothly and climb stairs.

Source

Nvidia Nvidia unveils the GeForce 256

Nvidia's GeForce 256 graphics processing unit was unveiled, with products from Creative promised for that fall.

Source

Paper LeNet-5 and the MNIST dataset

LeCun, Bottou, Bengio and Haffner described gradient-based learning for document recognition, including the LeNet-5 network used to read cheques.

Source

Paper Long short-term memory (LSTM)

Sepp Hochreiter and Jürgen Schmidhuber introduced LSTM, a recurrent network design that can retain information over long sequences.

Source

Event Deep Blue defeats Garry Kasparov

IBM's Deep Blue won a six-game match against world chess champion Garry Kasparov, the first match loss by a reigning champion to a computer under tournament conditions.

Source

Event Brewster Kahle founds the Internet Archive

Brewster Kahle started the Internet Archive in San Francisco, a nonprofit library that began crawling and saving the public Web, Netnews and downloadable software before they disappeared.

Source

Program Richard Wallace's A.L.I.C.E.

Richard Wallace released A.L.I.C.E., a pattern-matching chatbot written in AIML that later won the Loebner Prize three times.

Source

Paper Support-vector networks

Corinna Cortes and Vladimir Vapnik published support-vector machines, a classification method that dominated much of machine learning for the next decade.

Source

Program Tesauro's TD-Gammon

Gerald Tesauro of IBM built TD-Gammon, a neural network that learned backgammon by playing itself and reached near top human level.

Source

Event First Loebner Prize contest

The first Loebner Prize, an annual Turing-test style competition for chatbots, was held at the Computer Museum in Boston.

Source

Paper LeCun applies backpropagation to handwritten digits

Yann LeCun and colleagues at Bell Labs trained a convolutional neural network to read handwritten ZIP codes.

Source

Program Jabberwacky chatbot project begins

Rollo Carpenter began Jabberwacky, a chatbot meant to learn conversation from its users; it went online in 1997.

Source

Event The second AI winter

The market for specialised Lisp machines collapsed in 1987 and expert systems proved costly to maintain, leading to another drop in AI funding.

Source

Paper Rumelhart, Hinton and Williams popularise backpropagation

A paper in Nature showed how backpropagation of errors could train multi-layer neural networks to learn useful internal representations.

Source

Film The Terminator

James Cameron's film about Skynet, a defence AI that turns on humanity, became a common cultural reference for AI risk.

Source

Literature William Gibson's Neuromancer

Gibson's novel, featuring AIs named Wintermute and Neuromancer, popularised the idea of cyberspace and shaped cyberpunk fiction.

Source

Research The Cyc project begins

Douglas Lenat started Cyc, a long-running effort to hand-encode millions of common-sense facts and rules.

Source

Film Blade Runner

Ridley Scott's film, adapted from Philip K. Dick's novel, brought artificial humans called replicants to a wide audience.

Source

Paper Hopfield networks

Physicist John Hopfield described a recurrent neural network that stores memories as stable states, helping revive interest in neural networks.

Source

Event Japan's Fifth Generation Computer project

Japan's trade ministry launched a ten-year national programme to build logic-based computers for AI, prompting rival programmes in the U.S. and Europe.

Source

Paper Searle's Chinese room argument

Philosopher John Searle argued that a program manipulating symbols by rules would not understand language, even if its answers looked fluent.

Source

Program XCON expert system enters commercial use

Digital Equipment Corporation began using XCON (R1), a rule-based system that configured computer orders, a sign of the commercial expert systems boom.

Source

Event The first AI winter

After the Lighthill report and U.S. funding changes, money and interest in AI research fell sharply for several years.

Source

Event The Lighthill report

A review for the UK Science Research Council by James Lighthill judged that AI had failed to deliver on its promises, leading to major funding cuts in Britain.

Source

Program MYCIN expert system

Stanford's MYCIN used several hundred rules to recommend antibiotics for blood infections, performing comparably to specialists in tests though it was never used clinically.

Source

Program Colby's PARRY

Psychiatrist Kenneth Colby built PARRY, a program simulating a person with paranoid schizophrenia; it was later connected to ELIZA for a recorded machine-to-machine conversation.

Source

Invention Prolog programming language

Alain Colmerauer and Philippe Roussel created Prolog, a logic programming language widely used in AI and later chosen for Japan's Fifth Generation project.

Source

Program Winograd's SHRDLU

Terry Winograd's SHRDLU understood typed English commands about a simulated world of blocks and could answer questions about it.

Source

Paper McCarthy and Hayes describe the frame problem

John McCarthy and Patrick Hayes identified the frame problem: how a reasoning system can represent what stays unchanged when an action occurs.

Source

Paper Minsky and Papert's Perceptrons

Marvin Minsky and Seymour Papert analysed the limits of single-layer perceptrons, and the book is often linked to a decline in neural network research.

Source

Film 2001: A Space Odyssey and HAL 9000

Stanley Kubrick's film, written with Arthur C. Clarke, premiered with HAL 9000, a conversational computer that turns against the crew.

Source

Literature Philip K. Dick's Do Androids Dream of Electric Sheep?

Dick's novel follows a bounty hunter tracking androids nearly indistinguishable from humans, raising questions about empathy and identity.

Source

Event ALPAC report on machine translation

A U.S. advisory committee concluded that machine translation had not met expectations, leading to sharp cuts in funding for the work.

Source

Research Shakey the robot at SRI

SRI's Shakey was the first mobile robot to reason about its own actions, combining vision, planning and movement; the project produced the A* search algorithm.

Source

Program Weizenbaum's ELIZA

Joseph Weizenbaum published ELIZA, a program that held text conversations by matching patterns; its 'DOCTOR' script imitated a psychotherapist and many users attributed understanding to it.

Source

Program DENDRAL project begins

Edward Feigenbaum, Joshua Lederberg and colleagues at Stanford started DENDRAL, a program that inferred molecular structures from mass-spectrometry data and became the first expert system.

Source

Invention Unimate, the first industrial robot, goes to work

The Unimate robotic arm, developed by George Devol and Joseph Engelberger, began handling hot die-cast parts at a General Motors plant in New Jersey.

Source

Paper Arthur Samuel's checkers program and 'machine learning'

Arthur Samuel of IBM published his work on a checkers program that improved by playing itself, helping establish the term machine learning.

Source

Event McCarthy and Minsky start the MIT AI Project

McCarthy and Minsky founded the Artificial Intelligence Project at MIT, which grew into one of the field's leading laboratories.

Source

Invention John McCarthy creates Lisp

McCarthy designed Lisp, a programming language based on symbolic lists that became the main language of AI research for decades.

Source

Research Rosenblatt's Perceptron

Frank Rosenblatt published the perceptron, a trainable single-layer neural network, and demonstrated it for the U.S. Navy; the Mark I Perceptron hardware followed.

Source

Program General Problem Solver

Newell, Shaw and Simon began the General Problem Solver, a program meant to solve many kinds of formal problems using means-ends analysis.

Source

Event The Dartmouth Summer Research Project

About a dozen researchers met at Dartmouth College over roughly eight weeks; the workshop is widely treated as the founding event of AI as a field.

Source

Program Newell, Shaw and Simon's Logic Theorist

Allen Newell, Cliff Shaw and Herbert Simon presented the Logic Theorist, a program that proved theorems from Principia Mathematica, often called the first AI program.

Source

Paper Dartmouth proposal coins 'artificial intelligence'

John McCarthy, Marvin Minsky, Nathaniel Rochester and Claude Shannon proposed a summer research project on 'artificial intelligence', the first known use of the term.

Source

Event Georgetown–IBM machine translation demonstration

Georgetown University and IBM publicly demonstrated a computer translating more than sixty Russian sentences into English, the first public machine-translation demo.

Source

Program Strachey's draughts program

Christopher Strachey wrote one of the first game-playing programs, a draughts (checkers) player that ran on the Ferranti Mark 1.

Source

Invention Minsky and Edmonds build SNARC

Marvin Minsky and Dean Edmonds built SNARC, a machine of about 40 artificial neurons using vacuum tubes that simulated a rat learning a maze.

Source

Literature Asimov's I, Robot

Asimov collected his robot stories, built around the Three Laws, into the book I, Robot.

Source

Paper Turing's 'Computing Machinery and Intelligence'

Turing replaced the question 'Can machines think?' with the imitation game, now called the Turing test, and answered common objections to machine intelligence.

Source

Paper Shannon's 'Programming a Computer for Playing Chess'

Claude Shannon set out how a computer could play chess by evaluating positions and searching ahead, a template for game-playing programs.

Source

Invention Grey Walter's robot tortoises

Neurophysiologist William Grey Walter built small autonomous robots, Elmer and Elsie, that used light sensors to steer and find their charging station.

Source

Paper Hebb's The Organization of Behavior

Psychologist Donald Hebb proposed that connections between neurons strengthen when they are active together, a learning rule that influenced neural network research.

Source

Paper Turing's report 'Intelligent Machinery'

In an unpublished report for the National Physical Laboratory, Turing discussed machines that learn and early network-like 'unorganised machines'.

Source

Paper Norbert Wiener's Cybernetics

Norbert Wiener's book introduced cybernetics, the study of control and communication through feedback in animals and machines.

Source

Invention ENIAC is unveiled

The University of Pennsylvania publicly presented ENIAC, a programmable electronic general-purpose computer built for the U.S. Army.

Source

Paper Vannevar Bush's 'As We May Think' imagines the memex

In The Atlantic Monthly, Vannevar Bush described the memex, a desk that would store a person's documents on microfilm and link any two items into trails, an early vision of hypertext and personal information machines.

Source

Paper McCulloch and Pitts model the neuron

Warren McCulloch and Walter Pitts showed that networks of simple on/off model neurons could compute logical functions, the first mathematical model of a neural network.

Source

Literature Asimov's Three Laws of Robotics in 'Runaround'

Isaac Asimov's short story 'Runaround' set out the Three Laws of Robotics, rules meant to keep robots safe and obedient to humans.

Source

Paper Turing's 'On Computable Numbers'

Alan Turing described an abstract machine that manipulates symbols on a tape and showed that a universal machine could carry out any computation, a foundation of computer science.

Source

Film Fritz Lang's Metropolis

Lang's silent film, which features a robot made to impersonate a woman named Maria, premiered in Berlin and shaped later screen images of robots.

Source

Literature Karel Čapek's R.U.R. introduces the word 'robot'

Čapek's play about artificial workers who rebel against their makers premiered in Prague and gave the world the word robot.

Source

Invention Torres Quevedo's El Ajedrecista

Spanish engineer Leonardo Torres Quevedo built an electromechanical machine that could play a king-and-rook chess endgame against a human without help.

Source

Literature E. M. Forster's 'The Machine Stops'

Forster's short story imagines people living isolated lives entirely dependent on a global machine that eventually fails.

Source

Literature L'Ève future popularises the word 'android'

Auguste Villiers de l'Isle-Adam's novel features an artificial woman built by a fictional Thomas Edison and helped spread the term andréide (android).

Source

Literature Butler's Erewhon and 'The Book of the Machines'

Butler's novel Erewhon describes a society that has banned machines for fear they would develop consciousness.

Source

Literature Samuel Butler's 'Darwin among the Machines'

In a letter to a New Zealand newspaper, Samuel Butler suggested that machines might evolve and one day overtake humanity.

Source

Paper Boole's The Laws of Thought

George Boole published an algebra of logic in which reasoning is expressed with symbols and operations, the basis of the Boolean logic used in computers.

Source

Paper Ada Lovelace's notes on the Analytical Engine

Ada Lovelace translated Menabrea's article on the Analytical Engine and added notes, including a method to compute Bernoulli numbers and the remark that the machine could not originate anything on its own.

Source

Invention Babbage describes the Analytical Engine

Babbage designed a general-purpose mechanical computer with a store, a mill and punched-card input, though it was never completed.

Source

Invention Babbage proposes the Difference Engine

Charles Babbage announced plans for a machine to calculate mathematical tables automatically, the first of his mechanical computing designs.

Source

Literature Mary Shelley's Frankenstein

Mary Shelley's novel tells of a scientist who creates a living being and then abandons it, a lasting reference point for debates about creators and their creations.

Source

Literature E. T. A. Hoffmann's 'The Sandman'

Hoffmann's story follows a student who falls in love with Olimpia, a lifelike automaton, a theme later studied as the uncanny.

Source

Invention The Mechanical Turk chess automaton

Wolfgang von Kempelen presented a chess-playing automaton at the Habsburg court; it toured for decades before being revealed as a hidden human operator.

Source

Invention Jaquet-Droz automata: the Writer, the Draughtsman and the Musician

Swiss watchmakers Pierre Jaquet-Droz, his son Henri-Louis and Jean-Frédéric Leschot built three automata, including a boy who writes a programmable text with a quill.

Source

Invention Vaucanson's Digesting Duck

Jacques de Vaucanson exhibited a mechanical duck that appeared to eat and digest grain, one of the most famous automata of the 18th century.

Source

Literature Swift's 'Engine' in Gulliver's Travels

Jonathan Swift's satire describes a professor at Lagado whose engine of rotating word blocks produces books by random combination, an early parody of mechanical text generation.

Source

Idea Leibniz imagines a calculus of reasoning

Gottfried Wilhelm Leibniz's Dissertation on the Art of Combinations began his project of a universal symbolic language and a calculus ratiocinator in which disputes could be settled by calculation.

Source

Idea Hobbes: reasoning is reckoning

In Leviathan, Thomas Hobbes described reasoning as a form of calculation, adding and subtracting ideas, a view later linked to computational theories of mind.

Source

Idea Descartes asks whether machines could think

In the Discourse on the Method, René Descartes argued that a machine could imitate a body but could never use language flexibly or reason across all situations as a person does.

Source

Myth The golem of Prague legend

Jewish folklore tells of a golem, a clay figure animated through sacred words; the best-known version credits Rabbi Judah Loew of Prague with making one to protect the community.

Source

Invention Leonardo da Vinci's mechanical knight

Leonardo da Vinci sketched an armoured knight moved by pulleys and cables that could sit, stand and move its arms.

Source

Idea Ramon Llull's Ars Magna

The Majorcan philosopher Ramon Llull presented a method of combining basic concepts with rotating paper discs to generate statements mechanically, later cited as an ancestor of logical computation.

Source

Invention Al-Jazari's Book of Knowledge of Ingenious Mechanical Devices

The engineer Ismail al-Jazari completed an illustrated treatise describing water clocks and automata, including mechanical musicians and a serving figure.

Source

Invention Hero of Alexandria writes on automata

Hero of Alexandria described programmable automata and self-moving theatre devices driven by weights, ropes and pegged drums.

Source

Literature Ovid's Pygmalion: a statue brought to life

In the Metamorphoses, Ovid tells of the sculptor Pygmalion, whose ivory statue is brought to life, an early story of a made figure becoming a person.

Source

Myth Talos, the bronze guardian of Crete, in the Argonautica

Apollonius of Rhodes' Argonautica tells of Talos, a bronze giant made to guard Crete, who is defeated when the fluid in his single vein drains away.

Source

Myth Homer describes Hephaestus' golden attendants

The Iliad describes the smith god Hephaestus working with golden maidservants who can move, speak and think, one of the oldest images of artificial helpers.

Source