OpenAI Plans Purchase of Google TPUs to Diversify Beyond Nvidia Silicon
OpenAI is preparing to buy Google's tensor processing units, expanding its infrastructure beyond an exclusive reliance on Nvidia as compute costs climb.
Archive
Search
Microsoft Outlines Maia 300 Chip Details as Hyperscalers Expand In-House AI Silicon
Microsoft has detailed the architecture for its custom Maia 300 accelerator, setting plans to scale in-house silicon to multi-gigawatt deployments starting in 2027 as cloud providers increasingly seek to reduce AI inference costs.
