IBM is freeing the newest model of its mainframe {hardware} that comes with new updates supposed to boost up AI adoption.
The {hardware} and consulting corporate on Monday introduced IBM z17, the newest model of its mainframe laptop {hardware}. This absolutely encrypted mainframe is powered by means of an IBM Telum II processor and is designed for greater than 250 AI use circumstances, the corporate says, together with AI brokers and generative AI.
Mainframes would possibly appear to be outdated hat, however they’re utilized by 71% of Fortune 500 corporations lately, in step with one supply. In 2024, the mainframe marketplace was once price an estimated $5.3 billion, in keeping with consulting company Marketplace Analysis Long run.
The z17 can procedure 450 billion inference operations in an afternoon, a 50% build up over its predecessor, the IBM z16, which was once launched in 2022 and ran at the corporate’s authentic Tellum processor. The device is designed so to absolutely combine with different {hardware}, device, and open-source gear.
Tina Tarquinio, VP of product control and design for IBM Z, informed techmim that this mainframe improve has been within the works for 5 years — neatly prior to the present AI frenzy that began with the discharge of OpenAI’s ChatGPT in November 2022.
IBM spent greater than 2,000 analysis hours getting comments from over 100 shoppers because it constructed the z17, Tarquinio mentioned. She thinks it’s attention-grabbing to look that, now, 5 years later, the comments they were given aligned with the place the marketplace ended up heading.
“It’s been wild understanding that we’re introducing an AI accelerator, after which seeing, particularly within the later part of 2022, all the adjustments within the business relating to AI,” Tarquinio informed techmim. “It’s been actually thrilling. I feel the largest level has been [that] we don’t know what we don’t learn about what’s coming, proper? So the probabilities are actually limitless relating to what AI can lend a hand us do.”
The z17 is ready as much as adapt and accommodate the place the AI marketplace heads, Tarquinio mentioned. The mainframe will give a boost to 48 IBM Spyre AI accelerator chips upon unencumber, with the plan to deliver that quantity as much as 96 inside 365 days.
“We’re purposely construction in headroom,” Tarquinio mentioned. “We’re purposely construction in AI agility. In order new fashions are offered, [we’re] ensuring that we’ve constructed within the headroom for larger, better fashions — fashions that perhaps want extra native reminiscence to speak to one another. We’ve inbuilt that as a result of we realize it’s actually the means that can exchange, proper? The brand new fashions will come and pass.”
Tarquinio mentioned that one of the crucial highlights of this newest {hardware} — despite the fact that she joked it was once like being requested to pick out her favourite kid — is that the z17 is extra energy-efficient than its predecessor and supposedly competition, too.
“On-chip, we’re expanding the AI acceleration by means of seven and a part occasions, however that’s 5 and a part occasions much less calories than you would have to do, like, multi-model on any other form of accelerator or platform within the business,” Tarquinio mentioned.
The z17 mainframes will develop into typically to be had on June 8.
IBM,mainframe,ai infrastructure,AI {hardware}
Supply hyperlink
