Entity trail

Trainium

Source-backed findings, relationship evidence, citations, and briefing history from the public MindPattern archive.

Briefing refs

Findings

Edges

Sources

Corpus findings

2026-07-02 / reddit-researcherTwelveLabs Raises $100M Series B for 'Video Superintelligence,' Signs Multiyear AWS Trainium DealTwelveLabs raised a $100M Series B co-led by NEA and NAVER Ventures with Amazon participating, pushing total funding above $207M, to build 'video superintelligence' on its Marengo (scene/sound search) and Pegasus (video-to-text) models. Alongside the round, AWS signed a multiyear deal to host TwelveLabs workloads on Trainium chips and offer its models to Bedrock developers. The company is moving from video understanding into a full-stack agentic system for searching and operationalizing large video archives.
2026-05-23 / thought-leaders-researcherAnthropic in Early Talks to Rent Microsoft's Custom Maia 200 AI ChipsCNBC and Bloomberg report Anthropic is in early talks to rent Microsoft's second-generation Maia 200 chips, built on TSMC's 3nm process and designed for inference workloads. CEO Dario Amodei recently said the company has had 'difficulties with compute.' A deal would represent Anthropic's third custom silicon partnership — alongside AWS Trainium (10-year, $100B+ deal) and Google TPUs — and would be a win for Microsoft, which trails AWS and Google in supplying AI-specific chips to customers.
2026-05-03 / rss-researcherAmazon Q1 Earnings: AWS Grows 28% to $37.6B as Trainium Custom Silicon Bet Pays OffBen Thompson's Stratechery analysis of Amazon's earnings argues that the market's shift from training to inference and agents means Amazon's Trainium custom silicon bet is paying off. AWS grew 28% year-over-year to ~$37.6B, fueled by enterprise AI infrastructure demand. Thompson notes the irony: Amazon invested in custom chips while others focused on NVIDIA GPU access, and the inference-heavy agent era may favor exactly this approach.
2026-04-30 / skill-finderOpenAI-Amazon $50B Infrastructure Deal: 2GW Trainium Capacity, Exclusive Third-Party Cloud Distribution for OpenAI Frontier Enterprise Agent PlatformThe OpenAI-AWS partnership announced April 28 is backed by a $50B Amazon investment — the largest cloud AI infrastructure deal in history. The seven-year compute commitment includes 2 gigawatts of Trainium capacity and exclusive third-party cloud distribution rights for OpenAI Frontier, the enterprise agent platform. Microsoft's contract is restructured to a nonexclusive IP license through 2032. GeekWire notes OpenAI's models landed on Bedrock one day after Microsoft exclusivity formally ended. For builders: the multi-cloud era for OpenAI means you can now architect without cloud lock-in — evaluate GPT-5.5 on Bedrock with your existing AWS governance stack before committing to migration.
2026-04-30 / skill-finderOpenAI Ends Seven-Year Azure Exclusivity: GPT-5.5, Codex, and Managed Agents Launch on Amazon Bedrock in Limited PreviewOn April 28, OpenAI made GPT-5.5, GPT-5.4, Codex, and Bedrock Managed Agents available on AWS — ending nearly seven years of Microsoft Azure being the only hyperscaler hosting OpenAI's proprietary models. GPT-5.5 pricing: $5/$30 per 1M input/output tokens with 1M context; GPT-5.5-pro at $30/$180 for higher accuracy. Codex is accessible via CLI, desktop app, and VS Code extension with AWS credential auth. Backed by $50B Amazon investment and 2GW Trainium capacity commitment. For builders: multi-cloud OpenAI access means you can now use GPT-5.5 with existing AWS IAM, PrivateLink, and CloudTrail — no new security model required.
2026-04-30 / rss-researcherStratechery: Amazon's Trainium Bet Paying Off as Industry Shifts from Training to InferenceBen Thompson's Stratechery analysis argues that Amazon's bet on custom Trainium silicon is paying off as the AI industry shifts away from training toward inference and agents. The shift toward inference workloads — where cost-per-query matters more than peak training throughput — plays to AWS's strength in commodity-scale infrastructure. Thompson positions Trainium as AWS's answer to the question of whether you need NVIDIA GPUs for the inference-dominated future.
2026-04-30 / news-researcherAmazon Q1 2026: AWS Grows 28% to $37.6B (15-Quarter High), Plans Record $200B CapexAmazon beat Q1 estimates on April 29 with AWS revenue of $37.6B growing 28% — its fastest rate in 15 quarters — while total revenue reached $181.5B. Capital expenditures hit $44.2B for the quarter alone, with $200B planned for full-year 2026. Key AI commitments disclosed: OpenAI will consume ~2GW of Trainium capacity starting 2027, Anthropic secured up to 5GW of Trainium chips, and Meta signed for tens of millions of Graviton cores.
2026-04-28 / thought-leaders-researcherAmazon Finalizes $50B OpenAI Equity Investment + $100B Cloud Deal — OpenAI Hits $840B ValuationAmazon closed its $50B investment in OpenAI ($15B initial + $35B conditional), making it one of the largest single investments in tech history. The deal includes expanding an existing $38B cloud agreement by $100B over eight years with a 2GW Trainium chip commitment. Combined with NVIDIA and SoftBank funding, OpenAI reached an $840B valuation — directly triggering the Microsoft deal restructuring that ended exclusivity.
2026-04-24 / thought-leaders-researcherDwarkesh Patel Interview: Jensen Huang Reveals NVIDIA Brought Groq Into CUDA Stack for Premium Token MarketIn a 103-minute Dwarkesh Patel interview (April 15), Jensen Huang disclosed that NVIDIA has integrated Groq into the CUDA stack because the value of AI tokens has risen enough to create a premium market for faster response times, even at lower throughput. Huang also explained why Anthropic runs most compute on TPUs/Trainium: NVIDIA wasn't positioned to write the multi-billion dollar equity checks that Google and AWS made at Anthropic's founding. He predicted energy permitting — not silicon — as the binding constraint on AI buildout.
2026-04-21 / hn-researcherAnthropic and Amazon Expand Partnership: $25B Investment, 5GW of Compute, $100B AWS Commitment Over 10 YearsAnthropic announced a massive expansion with Amazon — $5B immediate investment plus up to $20B tied to commercial milestones, on top of $8B previously invested. Anthropic is committing $100B+ over ten years to AWS, securing up to 5GW of new compute capacity using Trainium2, Trainium3, and future chips. The deal reveals Anthropic's run-rate revenue has surpassed $30B (up from ~$9B at end of 2025), and the companies currently use over 1 million Trainium2 chips. International inference expansion in Asia and Europe included.
2026-04-16 / saas-disruption-researcherOpenAI Revenue Chief Touts Amazon Alliance in Internal Memo, Says Microsoft 'Limited Our Ability' to Reach ClientsOpenAI revenue chief Denise Dresser sent an internal memo on April 13 positioning the $50B Amazon partnership as the company's primary enterprise growth channel, explicitly stating that Microsoft has 'limited our ability' to reach enterprise clients. The partnership includes exclusive third-party cloud distribution through AWS Bedrock, a jointly-developed Stateful Runtime Environment, and an expanded $100B compute agreement over 8 years with 2GW of Trainium capacity. This is a structural fracture in the Microsoft-OpenAI relationship with direct implications for enterprise SaaS vendors building on either stack.
2026-04-10 / rss-researcherAmazon CEO Jassy Defends $200B Capex in Shareholder Letter — Values Custom Chip Business at $50B, Takes Aim at NvidiaAndy Jassy's 2026 shareholder letter reads as a competitive manifesto: Amazon's custom chip business (Graviton, Trainium, Nitro) generates $20B+ annualized revenue growing at triple-digit rates, and he values it at $50B if sold externally. Graviton now used by 98% of top 1,000 EC2 customers. The letter acknowledges Nvidia partnership while explicitly calling 'a new shift' away from Nvidia-only AI compute. Jassy also targeted Intel and Starlink while defending the $200B capex as backed by customer commitments.

Graph relationships

BUILT BY
Amazon
Trainium is Amazon's custom silicon for AI
Source finding
USES
Anthropic
Anthropic runs Claude at scale on Trainium chips
Source finding
BUILT BY
Amazon
Trainium is Amazon's custom AI chip
Source finding
USES
Anthropic
Anthropic runs Claude on over 1 million Trainium chips
Source finding
USES
Anthropic
Anthropic secured up to 5GW of Trainium chips.
Source finding
USES
OpenAI
OpenAI will consume 2GW of Trainium capacity starting 2027.
Source finding
RELEASED
Amazon
Amazon developed custom Trainium silicon for inference workloads.
Source finding
USES
OpenAI
OpenAI has adopted Trainium for AI inference
Source finding
COMPETES WITH
NVIDIA
Trainium competes with NVIDIA GPUs for AI inference cost-effectiveness.
Source finding
USES
Apple
Apple is testing Trainium for AI workloads
Source finding
USES
Apple
Apple has adopted Trainium for AI workloads
Source finding
USES
Anthropic -> Trainium
Anthropic runs Claude at scale on Trainium chips
Source finding

Source trail

Fortune/Bloomberg Stratechery analysis ArXiv paper The deal Bloomberg reports Amazon finalized a $50B equity investment in OpenAI Wall Street Journal reported Simon Willison published a detailed history Source OpenAI GeekWire VentureBeat

Graph sources

entity graphfindings textkg edgeskg entitiesnewsletter issues

Fetching from the wire…