Cloud roundup: Microsoft unveils its first AI chip

Microsoft launches its first AI chip

Tech giant Microsoft has launched its first AI chip, dubbed the Maia 100, according to Yahoo Finance. The Maia 100 can be used in cloud-based training and deploying AI models. It will also be part of Microsoft's upcoming Azure Maia AI accelerator series.

Microsoft will be testing the Maia 100 with ChatGPT developer OpenAI, and is also testing the chip internally with its Bing and Office AI products.

Microsoft also launched an Arm-based Azure Cobalt CPU, the Azure Cobalt 100, today.

The Azure Cobalt CPU server chip is "designed for cloud workloads and Microsoft already said it was 40 percent faster than the commercial Arm-based servers it uses today," reported The Verge.

By developing these chips in-house, "it could insulate Microsoft from becoming overly dependent on any one supplier, a vulnerability currently underscored by the industrywide scramble for Nvidia Corp.'s AI chips," explained Bloomberg's Dina Bass and Ian King.

The Maia and Azure Cobalt 100 chips will be deployed in select Microsoft data centers early next year.

Microsoft has also deployed a new software tool for customers to design their own AI assistants, according to Bloomberg.

AI blows up cloud budgets

For enterprises attempting to pull back on cloud spending, the growing use of and reliance on AI is making it harder to reduce their cloud costs. The global cloud market is forecast to reach nearly $679 billion next year, up about 20% from $564 billion this year, according to Gartner.

Related:Deutsche Telekom still favors telco DIY over public clouds

"Cloud has become essentially indispensable,” said Sid Nag, VP analyst at Gartner, in a statement. "However, that doesn’t mean cloud innovation can stop or even slow. The tables are turning for cloud providers as cloud models no longer drive business outcomes, but rather, business outcomes shape cloud models."

Nag added that organizations deploying generative AI services will rely on the public cloud due to the "scale of infrastructure required," but they will also seek out hyperscalers that address issues including cost, economics, sovereignty, privacy and sustainability.

Cloud infrastructure is key to supporting the development and use of AI models. While AI is increasing cloud spending, "some CIOs say it can also help them keep those costs in check and avoid 'cloud sprawl'," reported The Wall Street Journal. Essentially, AI could assist in analyzing cloud infrastructure options to identify the best way to budget for cloud spending.

City of Las Vegas deploys Juniper's Cloud Metro service

Related:Microsoft eyes Australia's AI and cloud market with A$5B investment

The City of Las Vegas is adding Juniper Cloud Metro services to build out its 5G network. The city said this is part of its goal to become a "model smart city by 2025."

Las Vegas is deploying Juniper's ACX7024 Cloud Metro Routers and EX4300 Switches to support its metro network. The Cloud Metro platform is managed by Juniper Mist Wired Assurance, an AI and cloud service that provides network visibility and automation.

Initially, the network will be deployed in the Las Vegas Innovation District, a proposed 20-acre area for tech companies and public spaces for events.

Las Vegas is home to more than 640,000 residents, visited by 42 million tourists annually and has 123 miles of fiber.