Microsoft makes data centre chips to speed AI operations and boost security
Designing custom chips can reduce Microsoft’s reliance on processors made by Intel and Nvidia
19 November 2024 - 16:13
byMax Cherney
Support our award-winning journalism. The Premium package (digital only) is R30 for the first month and thereafter you pay R129 p/m now ad-free for all subscribers.
A Microsoft logo is pictured on a store in the Manhattan borough of New York City, New York, US. File photo: CARLO ALLEGRI/REUTERS
San Francisco, US — Microsoft has designed two additional infrastructure chips for its data centres that will help speed artificial intelligence (AI) operations and increase data security, it said on Tuesday at its Ignite conference.
Microsoft has devoted significant resources to developing home-grown silicon for general-purpose applications and AI. Like rivals Amazon.com and Google, Microsoft’s engineers say there is a performance and price benefit to designing chips that are customised for its needs.
Designing custom chips can reduce Microsoft’s reliance on processors made by Intel and Nvidia.
Microsoft’s two new chips are designed to be installed deep within the company’s data centre infrastructure. One chip is designed to increase security and the other is for data processing.
The company makes the effort to design an array of data centre processors because it aims to “optimise every layer of infrastructure” and ensures that Microsoft’s data centres crunch information at the speed AI requires, said Rani Borkar, corporate vice-president at Azure Hardware Systems and Infrastructure
Engineers will install the new security chip, called the Azure Integrated HSM, in every new server destined for a data centre beginning next year. The chip aims to keep crucial encryption and other security data inside the security module.
The data processing unit, or DPU, aims to move multiple components of a server into a single chip that is focused on cloud storage data. The company said it could run these specific tasks at three times less power and four times the performance compared with its current hardware.
Microsoft also announced a new version of a cooling system for data centre servers that relies on liquid to reduce the temperature of nearby components. The cooling unit can be used to support large-scale AI systems.
Support our award-winning journalism. The Premium package (digital only) is R30 for the first month and thereafter you pay R129 p/m now ad-free for all subscribers.
Microsoft makes data centre chips to speed AI operations and boost security
Designing custom chips can reduce Microsoft’s reliance on processors made by Intel and Nvidia
San Francisco, US — Microsoft has designed two additional infrastructure chips for its data centres that will help speed artificial intelligence (AI) operations and increase data security, it said on Tuesday at its Ignite conference.
Microsoft has devoted significant resources to developing home-grown silicon for general-purpose applications and AI. Like rivals Amazon.com and Google, Microsoft’s engineers say there is a performance and price benefit to designing chips that are customised for its needs.
Designing custom chips can reduce Microsoft’s reliance on processors made by Intel and Nvidia.
Microsoft’s two new chips are designed to be installed deep within the company’s data centre infrastructure. One chip is designed to increase security and the other is for data processing.
The company makes the effort to design an array of data centre processors because it aims to “optimise every layer of infrastructure” and ensures that Microsoft’s data centres crunch information at the speed AI requires, said Rani Borkar, corporate vice-president at Azure Hardware Systems and Infrastructure
Engineers will install the new security chip, called the Azure Integrated HSM, in every new server destined for a data centre beginning next year. The chip aims to keep crucial encryption and other security data inside the security module.
The data processing unit, or DPU, aims to move multiple components of a server into a single chip that is focused on cloud storage data. The company said it could run these specific tasks at three times less power and four times the performance compared with its current hardware.
Microsoft also announced a new version of a cooling system for data centre servers that relies on liquid to reduce the temperature of nearby components. The cooling unit can be used to support large-scale AI systems.
Reuters
WAYNE EPSTEIN: How diversity and AI drive business success
DUNCAN MCLEOD: Voice computing will be the next big thing
Tesla soars past $1-trillion valuation on Republican victory
Would you like to comment on this article?
Sign up (it's quick and free) or sign in now.
Please read our Comment Policy before commenting.
Most Read
Related Articles
Sweden’s Klarna files for long-awaited IPO in the US
Alliance aims to improve the tracking of medicines using 2D bar codes
US orders Taiwan AI chipmaker to halt shipments to China
Nvidia to bump Intel off Dow Jones industrial average
Published by Arena Holdings and distributed with the Financial Mail on the last Thursday of every month except December and January.