At its Ignite conference in Seattle, Microsoft introduced two new chips – the Maia 100 artificial intelligence chip and the Cobalt 100 Arm chip. These chips are designed to address specific computing needs and could potentially compete with industry leaders such as Nvidia and Intel.
The Maia 100 AI chip is poised to rival Nvidia’s highly sought-after AI graphics processing units (GPUs). Microsoft aims to leverage this chip for AI computing tasks, with potential applications in its Bing search engine’s AI chatbot, GitHub Copilot coding assistant, and the GPT-3.5-Turbo language model developed by Microsoft-backed OpenAI.
Indeed, the GPT-3.5-Turbo model within OpenAI’s ChatGPT assistant gained rapid popularity shortly after its release last year. This surge in popularity prompted companies to swiftly integrate comparable chat capabilities into their software, leading to an increased demand for Graphics Processing Units (GPUs) due to the computational requirements of such AI models.
Tekedia Mini-MBA edition 16 (Feb 10 – May 3, 2025) opens registrations; register today for early bird discounts.
Tekedia AI in Business Masterclass opens registrations here.
Join Tekedia Capital Syndicate and invest in Africa’s finest startups here.
“We’ve been working across the board and [with] all of our different suppliers to help improve our supply position and support many of our customers and the demand that they’ve put in front of us,” CNBC quoted Colette Kress, Nvidia’s finance chief, as saying at an Evercore conference in New York in September.
The Cobalt 100 Arm chip, on the other hand, is targeted at general computing tasks and could compete with Intel processors. Virtual-machine instances running on Cobalt chips are expected to be commercially available through Microsoft’s Azure cloud in 2024.
Microsoft’s move to introduce these chips aligns with the trend among technology companies, including Alibaba, Amazon, and Google, to provide clients with diverse options for cloud infrastructure.
This development comes at a time when cash-rich tech companies are expanding their offerings in the cloud infrastructure space. Microsoft, with a substantial cash reserve of about $144 billion, has a 21.5% market share in the cloud as of 2022, trailing only Amazon.
While the Maia 100 is currently being tested for various applications, including Bing, GitHub Copilot, and GPT-3.5-Turbo, the timeline for its release was not specified. However, the Cobalt processors are expected to be available for commercial use in 2024 through Microsoft’s Azure cloud.
In recent years, specialized AI chips have become crucial in meeting the demand for AI computing, especially during periods of GPU shortages. Microsoft’s approach involves developing these chips for specific tasks based on customer feedback.
Unlike some competitors, Microsoft and its cloud computing peers are not planning to sell servers containing their chips directly to companies. Instead, the chips will be utilized within the companies’ cloud infrastructure offerings.
With Google, Amazon, and others having previously introduced their own specialized chips for AI and general computing, Microsoft aims to enhance performance, efficiency, and cost-effectiveness within cloud environments with these chips.
Microsoft’s introduction of the Maia 100 and Cobalt 100 Arm chips underscores how heated the race has become in the chip industry, and the company’s plan to tag along.
Microsoft could experience quicker adoption of Cobalt processors compared to the Maia AI chips, drawing from Amazon’s experience as a reference point. Microsoft has been conducting tests of its Teams app and Azure SQL Database service on Cobalt processors.
According to Microsoft’s statements, these tests have demonstrated a performance improvement of 40% compared to Azure’s current Arm-based chips, which are supplied by the startup Ampere. This promising performance boost could potentially pave the way for the faster integration and adoption of Cobalt processors within Microsoft’s infrastructure.