Microsoft's first in-house chips are centered around AI

Microsoft is one of the biggest companies in the entire world, yet it’s shied away from developing its own in-house chips. Its phones, consoles, and other components use chips manufactured by other companies. However, the company just announced that this is about to change. According to a new report, Microsoft is developing its own in-house chips, and, as you would guess, they’re centered around AI.

Microsoft, along with the rest of the world, has been jumping on the AI craze as of late. Companies are pouring billions of dollars into developing smarter and more intuitive AI. Heck, Microsoft basically adopted OpenAI as a child after chat GPT launched. So, it should come as no surprise that Microsoft is developing its own silicon for AI.

Microsoft is developing its chips to help boost its AI capabilities

These new chips from Microsoft are targeted at cloud-based artificial intelligence. They’re named the Cobalt 100 chip and Maia 100 chip and the company is going to use these chips for different purposes.

The Maya 100 chip will be used to train AI models. This is similar to how people have been buying Nvidia’s h100 GPUs en masse to train their AI models. Just recently, Elon Musk bought a ton of Nvidia GPUs to train the AI models for his companies. As for the Cobalt 100 chip, this ship will be used to power the company’s Azure Cloud platform.

For the time being, Microsoft has had to rely on other companies like Nvidia for chips to power its AI ventures. However, with its own in-house chips, Microsoft will be able to cut costs. It’s always cheaper to manufacture your own components than to buy from third-party sellers.

Right now, the company is still working on manufacturing these ships, so you won’t see anything soon. However, the report states that these chips should be coming out sometime in 2024. This means that Microsoft has been working on them for some time.

This shows how extremely committed Microsoft is to AI technology at this point. We expect it to be implemented into more of the company’s services. However, we don’t know exactly how many.