Microsoft Preparing to Deploy Its Own Chips at Multi-Gigawatt Capacity in Data Centers

Microsoft is reducing its reliance on Nvidia by increasingly using its own AI chips. The company is preparing to install the latest generation of its custom chips in data centers. Its goal is to run its AI software as much as possible using s…

According to CNBC, Microsoft is reducing its dependence on Nvidia by increasing the use of its proprietary artificial intelligence chips. Microsoft the heavy cost of bypassing the Strait of Hormuz; The hidden reality of the return of Persian Gulf oil exports at a glance, Persian Gulf oil exports returned to about 16.5 million barrels per day in September, and is Bloom Energy's stock still worth buying after growing 15 times? Bloom Energy shares have grown by more than 1,400% in the last five years, and only since October last year, Stornaras said about the gradual approach of the European Central Bank in raising interest rates.

Meta Platforms Company is on the verge of publishing its third quarter financial report; An event that can once again increase the shares of this giant is the reaction of the German chancellor to the American-Russian agreement: Sanctions should be tougher. German Chancellor Friedrich Mertes said about the American-Russian agreement that the "Diesel" agreement is out of the question for Germany and that the sanctions should be intensified by subscribing to the YouTube Times newsletter.

According to CNBC, Microsoft is reducing its dependence on Nvidia by increasing the use of its proprietary artificial intelligence chips.

Microsoft is preparing to install its latest generation of proprietary chips in data centers. The company's goal is to run its AI software as much as possible using hardware it has designed and manufactured.

Microsoft announced on Monday that it plans to deploy several gigawatts of Maia-series chips in its data centers and server fleets. The release of the new generation of these chips, namely Maia 300, will also begin next year.

These semiconductors are designed to perform AI inference tasks; A process in which an artificial intelligence model is used to perform tasks such as summarizing a conference call or providing an answer to a research question.