EXPLORINEER
en

UK — The Register

PyTorch 2.15 tackles the 'one bad GPU' problem

AI

The PyTorch 2.15 release includes a rewritten backend for Nvidia's NCCL library, aimed at improving fault diagnosis and recovery when GPUs work together. The new backend allows only one communicator per process group, can exclude slow or crashed GPUs with a third-party orchestrator, and adds a "flight recorder" that captures operations for inspection after failures. The legacy backend remains as a rollback option in 2.15 and is slated for removal in the 2.16 release, planned for early next year.

Ad

Add 150 words. Learn 10 a day.

Dicter word books send only the words you learn to your review notes.

More in this topic

AITechCrunch

Meta introduces Spaces, a new way to collaborate with your Muse agent

Meta has launched Spaces, which it calls "custom experiences," as a new way to communicate and collaborate with its Muse agents. Spaces are topic-focused conversations that appear in the interface's sidebar, where Muse works toward goals such as planning a vacation or tracking holiday gifts. Meta says Spaces can connect to third-party apps and update in the background, and Spaces follows Muse Voice, which was announced the previous Thursday.

AIVnExpress

The Risk of Getting Trapped When Following AI Directions

Travelers increasingly rely on AI to plan outdoor trips, leading to dangerous incidents that require emergency rescues. Wired reported that three climbers had to call for rescue on Mount Shasta in California after overestimating their fitness, carrying too little food, and trusting Google's Gemini chatbot, while experts cite a study finding chatbot travel planning accurate only 10% of the time in one case.

AITechCrunch

After blocking Muse, Amazon now lets users text Alexa to help them shop

Amazon has launched a feature that lets select U.S. customers chat with its Alexa AI agent over text messages to search for items, add them to cart, and make purchases. The rollout comes months after Amazon blocked Meta's Muse from querying its site, and the company plans to expand access in the coming weeks ahead of the holiday shopping season.

AITechCrunch

Underdog shows off a 32MB speech model in new demo

Underdog has released a browser-based game built with its Woof voice model, which is only 32MB and has 31 million parameters. The company says the model runs on most devices without an internet connection, and it is open for anyone to download. The demo follows Underdog's Bark speech-to-text model, a 167 MB system for transcription and dictation designed for edge devices.