FILE PHOTO: The IBM logo is seen during the Viva Technology conference dedicated to innovation and startups at Porte de Versailles exhibition center in Paris, France, June 12, 2025. REUTERS/Benoit Tessier/File Photo
SAN FRANCISCO (Reuters) -International Business Machines on Tuesday announced a new line of data center chips and servers that it says will be more power-efficient than rivals and will simplify the process of rolling out artificial intelligencein business operations.
IBM introduced its new Power11 chips on Tuesday, marking its first major update to its "Power" line of chips since 2020.
These chips have traditionally vied against offerings from Intel and Advanced Micro Devices in data centers, particularly in specialized sectors such as financial services, manufacturing and healthcare.
Like Nvidia's AI servers, IBM's Power systems are an integrated package of chips and software.
Tom McPherson, general manager of Power systems at IBM, said the Armonk, New York-based company used that tight coupling to focus on reliability and security.
The Power11 systems, available from July 25, will not need any planned downtime for software updates, and their unplanned downtime each year averages just over 30 seconds.
They are also designed to detect and respond within a minute to a ransomwareattack - where hackers encrypt data and then try to extract a ransom in exchange for the keys, IBM said.
In the fourth quarter of this year, IBM plans to integrate Power11 with Spyre, its AI chip introduced last year.
McPherson said IBM does not aim to compete with Nvidia in helping create and train AI systems, but is instead focused on simplifying AI deployment for inference, the process of putting an AI system to work in speeding up a business task.
"We can integrate AI capabilities seamlessly into this for inference acceleration and help their business process improvements," McPherson said in an interviewlast week referring to work with early customers.
"It's not going to have all the horsepower for training or anything, but it's going to have really good inferencing capabilities that are simple to integrate."
(Reporting by Stephen Nellis in San Francisco; Editing by Himani Sarkar)
