Tech
DeepSeek plans to train its V5 model on Huawei Ascend 950 chips
DeepSeek plans to use at least 160,000 Huawei Ascend 950DT chips to train its V5 model in a new data center in Ulanqab, Inner...
Tech
Google and University of Hong Kong researchers want AI agents to show their work
Google and the University of Hong Kong are introducing VeriCUA, a verifier that checks the actual state of applications and files before accepting an...
Tech
Hyperscalers reportedly deploying around 1,000 NVL72 racks a week
Nvidia is now shipping more than 1,000 MGX racks per week, while hyperscalers are reportedly deploying about 1,000 NVL72 racks, equivalent to 72,000 Blackwell...
Tech
Nvidia AI Cluster Runtime packages Dynamo into tested Kubernetes recipes
Nvidia AI Cluster Runtime (AICR) combines the distributed inference framework Dynamo and its Kubernetes dependencies with recipes whose component versions have been validated together....
Tech
OpenAI’s Pando cuts Git push latency by about 80% for its monorepo
OpenAI developed Pando, a custom Git server that cuts the median latency for pushing code to the company’s monorepo by about 80%, nearly a...
Tech
Y Combinator CEO Garry Tan says he spends up to $200,000 a year on AI tokens
Garry Tan, president and CEO of Y Combinator, estimates spending between $100,000 and $200,000 a year on AI tokens, the usage credits charged when...
Tech
Salesforce agrees to acquire AI startup Cresta for about $2.7 billion
Salesforce has reached a definitive agreement to acquire Cresta, a startup specializing in voice-based customer service automation, for approximately $2.7 billion. Cresta’s platform enables...
Tech
Taurus Adds Stellar Smart-Contract Assets to Institutional Custody and Issuance
Taurus announced on October 29 that its Taurus-PROTECT and Taurus-CAPITAL platforms now support Stellar assets based on smart contracts. The development extends the company’s...