Intel’s CEO Pat Gelsinger said at the company’s annual Investor day that it plans to ship more than 4 million graphics processing units in 2022. These GPUs will span across chips for workstations, laptops and desktops.
The first plan in this regard is to release graphic cards for laptops in the first quarter of 2022, under the brand name Intel Arc Graphics, also known as Alchemist. This will be followed by graphics cards for desktops in the second quarter and workstation specific GPUs in the third.
Intel has partnered with companies such as Acer, Asus, Dell, Samsung, HP and Lenovo to deliver these products. The company said it has software partnerships to the tune of a 100 who are engaged in integrating their applications into Intel’s platforms.
Intel also made its intentions to target hardcore gamers clear with two high-powered GPUs slated for release in the 2023-2024 time period. These GPUs are code-named Battlemage and Celestial, with Celestial being slated for a late 2024 launch.
Also read: https://www.techcircle.in/2022/02/14/intel-unveils-crypto-chips-dorsey-s-block-among-first-clients
“Architecture work has begun on Celestial, a product that will address the ultra-enthusiast segment,” said Raja Koduri head of Intel’s Accelerated Compute Systems and Graphics Group.
For the broader picture, apart from promising that Intel will double its revenue and increase shareholder value by as much as four times over the decade. Gelsinger added that the chip maker also could become the US stock market’s “next great growth story”.
The company plans to increase its annual revenue growth to reach the 10%-12% range by 2025, its revenue had fallen by 4% in 2021.
Other key announcements at investor day included a new roadmap for Xeon Central processing units, new CPU and graphic processing unit hybrid products, and a new GPU aimed solely for media processing.
In the supercomputer spectrum, Intel announced a new architecture called Falcon Shores, expected to go on sale in 2024, and will “provide a scalable architecture for all high-performance computing workloads,” said Koduri.