- Microsoft unveils Maia 200 AI hardware
- Maia 200 reportedly offers more performance and efficiency than rivals AWS and GCP
- Microsoft will use it to help improve Copilot internally, but it will also be available to customers.
Microsoft has revealed Maia 200, its “next big milestone” in supporting the next generation of artificial intelligence and inference technology.
The company’s new hardware, the successor to the Maia 100, will “dramatically change the economics of AI at scale”, offering a significant improvement in terms of performance and efficiency to stake a claim in the market.
The launch will also look to boost Microsoft Azure as a great place to run AI models faster and more efficiently, as it looks to take on big rivals Amazon Web Services and Google Cloud.
Microsoft Maia 200
Microsoft says the Maia 200 contains more than 100 billion transistors built on the TSMC 3nm process with native FP8/FP4 tension cores, a redesigned memory system with 216GB HBM3e at 7TB/s, and 272MB of on-chip SRAM.
All of this contributes to the ability to deliver over 10 PFLOPS at 4-bit precision (FP4) and around 5 PFLOPS at 8-bit performance (FP8), easily enough to run even the largest AI models available today and with room to grow as the technology evolves.
Microsoft says Maia 200 delivers 3x the FP4 performance of 3rd-generation Amazon Trainium hardware and higher FP8 performance than Google’s 7th-generation TPU, making it the company’s most efficient inference system yet.
And thanks to its optimized design, which sees the memory subsystem focused on narrow-precision data types, a specialized DMA engine, integrated SRAM, and a specialized NoC fabric for high-bandwidth data movement, the Maia 200 is capable of maintaining more of a model’s weights and data locally, meaning fewer devices are required to run a model.
Microsoft is already using the new hardware to power its AI workloads in Microsoft Foundry and Microsoft 365 Copilot, with greater availability to customers coming soon.
It is also deploying Maia 200 in its US Central data center region, with more deployments coming soon to its US West 3 data center region, near Phoenix, Arizona, and other nearby regions.
For those who want an early look, Microsoft is inviting academics, developers, cutting-edge AI labs, and open source model project contributors to register now to preview the new Maia 200 software development kit (SDK).
Follow TechRadar on Google News and add us as a preferred source to receive news, reviews and opinions from our experts in your feeds. Be sure to click the Follow button!
And of course you can also follow TechRadar on TikTok for news, reviews, unboxings in video form and receive regular updates from us on WhatsApp also.
