Microsoft is rolling out its second-generation AI chip, the Maia 200, as the corporate seeks to enhance effectivity throughout its cloud and AI companies whereas lowering dependence on Nvidia {hardware}.
The processor is being manufactured by Taiwan Semiconductor Manufacturing and deployed to Microsoft information facilities in Iowa, with further rollouts deliberate for the Phoenix space.
A number of the first chips will likely be allotted to Microsoft’s superintelligence group to generate information for future AI mannequin improvement, cloud and AI chief Scott Guthrie mentioned in a weblog submit.
The Maia 200 may even energy Copilot for enterprise prospects and assist AI fashions, together with these from OpenAI, provided via Microsoft’s Azure cloud platform.
Microsoft started creating its personal chips years after Amazon and Alphabet launched inside silicon efforts aimed toward reducing prices and enhancing efficiency.
Shortage and excessive costs of Nvidia’s newest processors have intensified the race amongst cloud suppliers to safe various sources of computing energy.
The corporate mentioned Maia 200 delivers stronger efficiency on some AI workloads than comparable chips from Amazon Internet Providers and Google and is probably the most environment friendly inference system Microsoft has deployed up to now.
Microsoft is already designing a successor, the Maia 300, and retains entry to OpenAI’s early-stage chip designs beneath its strategic partnership.
Microsoft shares rose almost 2% Monday as buyers positioned forward of the corporate’s fiscal second-quarter earnings launch on Wednesday. The software program maker’s market capitalization hovered close to $3.85 trillion.
In the meantime, Nvidia shares have been buying and selling about 0.5% decrease on the day because the chipmaker introduced a $2 billion funding in CoreWeave.
