Amazon is making a direct play against Nvidia's software dominance with TorchNeuron, a new tool that lets developers run existing PyTorch code on AWS Trainium chips. The setup requires just a single configuration change, removing the need to rewrite code for different hardware. Amazon says about 85% of AI applications use PyTorch, and it plans to open-source TorchNeuron at the PyTorch Conference in San Jose on October 20. Over 30 companies tested the tool during a private beta phase, and Amazon is also contributing changes to PyTorch itself to reduce its reliance on Nvidia.
Key Takeaways
- Amazon released TorchNeuron, enabling PyTorch developers to run code on Trainium chips with one configuration change
- Amazon plans to open-source TorchNeuron at the PyTorch Conference on October 20
- Over 30 companies tested TorchNeuron during its private beta phase
- Amazon is contributing changes to PyTorch to make it less tied to Nvidia
- About 85% of AI applications use PyTorch, according to Amazon
Amazon lets PyTorch developers use Trainium AI chips without code changes
Amazon released TorchNeuron, a tool that lets developers run existing PyTorch code on AWS Trainium chips without rewriting anything. The tool requires only a single configuration change to switch hardware. Amazon will open-source TorchNeuron at the PyTorch Conference on October 20. Startups and enterprises in robotics, healthcare, and finance are already building on it.
Amazon launches TorchNeuron to challenge Nvidia's AI software dominance
Amazon introduced TorchNeuron, a software tool that runs PyTorch code on its Trainium chips. The tool aims to make switching from Nvidia as simple as changing a setting. Amazon plans to release TorchNeuron as open-source software at the PyTorch Conference in San Jose. Over 30 companies tested TorchNeuron during its private beta phase.
Amazon unveils TorchNeuron in latest bid to rival Nvidia
Amazon Web Services released TorchNeuron, a tool that lets AI apps built with PyTorch run on Trainium chips with one configuration change. Amazon says 85% of AI apps use PyTorch, and TorchNeuron removes the need to rewrite code for different hardware. Amazon is also contributing changes to PyTorch itself to make it less tied to Nvidia. The move could reshape the AI chip market.
Intuit to buy Sequence to expand AI-powered mid-market software
Intuit announced an agreement to acquire Sequence, an AI-native revenue management platform. The deal will strengthen Intuit Enterprise Suite and QuickBooks Online Advanced with automated billing and pricing tools. Sequence replaces manual processes with AI-powered workflows for complex revenue models. The acquisition could help Intuit compete with Oracle and SAP in the mid-market.
Thomson Reuters strengthens growth through AI innovation and portfolio changes
Thomson Reuters reported 8% organic revenue growth in Q2 2026 and raised its full-year outlook to about 8%. The company is advancing AI through its Westlaw Advantage platform and the CoCounsel Legal beta. It also agreed to sell a majority stake in its Global Print business to KKR for roughly $500 million. These moves aim to sharpen focus on AI-powered professional information services.
Swap test shows GPT-5.5 and Kimi lead AI model comparison
AI Engineer tested four models, GPT-5.5, Kimi K2.6, DeepSeek V4 and Mistral Large 3, on a shared Microsoft Foundry project. Only two passed a 13-E test covering grounding and vision tasks. GPT-5.5 and Kimi K2.6 correctly identified animals and prices in multimodal images, while Mistral Large 3 made errors. GPT-5.5 also performed best in tool calling for calendar events and calculations.
AI predicts cancer death risk from facial images
Researchers trained an algorithm to predict cancer patients' risk of dying using just an image of their face. For some patients, the AI predictions were shockingly accurate. The study explores whether facial data alone can signal health outcomes. The results raise both hope and concern about AI in medicine.
AI stocks face earnings pressure as investment returns worry investors
Tech earnings this week are under scrutiny as investors worry about returns on big AI investments. CoreWeave stock sold off last week, adding to the jitters. Companies reporting include SpaceX, Meta Platforms, Texas Instruments and Lam Research. Investors are watching management commentary closely for signs of AI spending sustainability.
AI Glasses Forum set for Nov. 26 as China sales surge
The 10th AI/AR Smart Glasses Technology Industry Forum will be held on Nov. 26 covering optical waveguides, Micro-LED displays, AI agents and chip-sensor integration. China's online smart glasses retail sales hit CNY 560 million in Q1 2026 with 282,000 units sold. IDC forecasts China shipments will grow from 2.46 million units in 2025 to 4.508 million in 2026. Fifty-six AI glasses stocks drew over 100 institutional research visits each in the past year.
AI stocks rally as chip stocks lead Nasdaq rebound
The Nasdaq rose 1.4% on Monday as investors rotated back into chip stocks after last week's selloff. Concerns about the sustainability of the AI boom had caused several tech IPOs to be delayed or canceled. Oil prices fell and gold prices rose, helping drive the rally. Some analysts believe the market drop is a normal correction and the AI boom will continue long term.
Atlassian launches AI agent protocol with human accountability
Atlassian CEO Mike Cannon-Brookes introduced the Agentic Multiplayer Protocol on Oct. 6. The system gives AI agents visible profiles and tracks their edits like a coworker's. It aims to handle workplace concerns about permissions, security, and leaks. Cannon-Brookes stressed that a human must take responsibility for agent mistakes. He also highlighted Atlassian's Australian roots and global growth.
Meta faces new privacy scrutiny over AI smart glasses
Meta Platforms is facing fresh legal scrutiny over its AI-enabled smart glasses. The devices capture continuous audio, photos, and video data. Privacy advocates question if always-listening features follow consent rules. Regulators are assessing how wiretap laws apply to Meta's real-time AI recording. The debate affects Meta's core business as a large interactive media group.
Geotab deploys AI to 3,400 employees in one day
Geotab CEO Neil Cawse says companies that delay AI adoption will fall behind. His firm rolled out a custom AI called GIA to 3,400 employees in one day. Within seven days, 97% of employees used it daily. Cawse shared six lessons including simple onboarding, clear communication, and setting guardrails. He urged leaders to act quickly rather than waiting for a perfect plan.
Sources
- Amazon opens its AWS Trainium AI chips to every PyTorch developer, no rewrite needed
- Amazon challenges Nvidia’s software advantage with new tool for its AI chips
- Amazon is taking its biggest shot at Nvidia yet with new TorchNeuron dev tool
- Intuit to Acquire Sequence to Boost AI-Powered Mid-Market Offerings
- Thomson Reuters' AI Push Gains Steam: Is TRI Poised for Growth?
- Swap test shows GPT-5.5 and Kimi ace count
- Can AI predict whether you’ll die from cancer from just an image of your face?
- AI Stocks Face Earnings Gut Check As Spending Jitters Mount
- AI Glasses Forum Set for Nov. 26 as China's Q1 Online Retail Sales Hit CNY 560 Million, Shipments Seen Up 80%
- AI stocks rally ahead of Meta earnings
- Atlassian’s Mike Cannon-Brookes: AI agents need a name tag—and a human to blame for their mistakes
- Meta (META) Faces Fresh Privacy Scrutiny Over AI Smart Glasses
- Companies waiting on AI will fall behind
Comments
Please log in to post a comment.