Today we are announcing our biggest bet yet on our own silicon.

Microsoft is moving its own inference onto in-house silicon at multi-gigawatt scale from next calendar year, claiming large tokens-per-watt gains. That shapes inference cost and capacity for products built on its platforms.

• inference — Running a trained model to get answers — the phase where AI is actually used, as opposed to trained.
• token — The chunk of text a model reads and writes in — roughly three-quarters of a word — and the unit AI usage is billed in.

• articleNadella: treat frontier models as insider risks and contain them
• postThe Reverse Information Paradox

• postOpenAI's first inference chip posts results, deploys by year-endOpenAI
• articleSpaceX 10GW in 2027 – Why It’s Real, Will Drive $300B ARR for SpaceX, and Why Microsoft Will Be the Largest OfftakerJeremie Eliahou Ontiveros
• postLocal inference crosses over: the M5 Ultra and the 1,000x claimAlex Cheema

Checking sign-in…

Loading comments…