One of the primary reasons for Microsoft’s foray into chip-making is to reduce costs, which Athena could possibly slash by a third in comparison to Nvidia
Microsoft has announced its plan to develop its own AI chips, which will be used to power its Azure cloud platform and other products. This move will bring Microsoft into the growing club of tech companies that have started building their own custom chips to handle AI workloads.
According to Microsoft’s Chief Technology Officer, Kevin Scott, the company has already started to make progress in developing its own AI chips. He said, “We are going to build some of the world’s most advanced AI supercomputers, AI chips and infrastructure.”
Scott also mentioned that Microsoft is collaborating with researchers from several universities to develop these AI chips. He stated, “We’re working with a number of different academic institutions to do advanced research around these new types of chips.”
Microsoft is not the first company to enter the AI chip market. Other tech giants such as Google, Amazon, and Apple have already developed their own chips to support AI workloads. Google has developed its own Tensor Processing Unit (TPU), while Amazon has its own Inferentia chips, and Apple has its Neural Engine.
According to Gartner, AI chip revenue is expected to reach $34.3 billion by 2023, up from $4.6 billion in 2018. This highlights the significant growth potential for companies that develop their own AI chips.
In conclusion, Microsoft’s entry into the AI chip market will enable the company to provide better AI capabilities to its customers and stay competitive in the rapidly evolving tech landscape. As Kevin Scott put it, “We’re going to need to have a lot of this foundational technology in place to really build these amazing AI solutions that we want to build.”
By The Impactlab