Intel Corporation, the world’s largest chip maker, is gearing up to provide an update on its Artificial Intelligence (AI) and Data Center plans. The company will hold an investor event soon where it is expected to present the latest on its strategies for these two critical segments.
Intel betting big on AI technology
Intel has been betting big on AI technology as it looks to capitalize on the growing interest of investors in this field. The company has been highlighting its processing capabilities, led by the Xeon range, to attract investment into AI.
The recent announcement of investing $20 billion in two major manufacturing plants in Arizona was aimed at keeping the company competitive in semiconductor manufacturing. Despite the challenges faced by Intel, CEO Pat Gelsinger affirmed that the company would continue with its focus on chip manufacturing.
Mixed news for Intel’s Data Center and AI group
While Intel’s Data Center and AI group suffered a 15% drop in revenue year-on-year to $19.2 billion, there are some positive developments for this segment. The Sapphire Rapids CPUs are finally shipping in high volumes this quarter as part of the 4 Generation Xeon Scalable lineup. Its CPU successors are also coming up quickly.
However, there have been notable stumbles in 2021/2022 with the unexpected cancellation of Rialto Bridge, which would have been Intel’s next Data Center GPU Max product.
Intel optimistic about FPGA and dedicated AI silicon products
Intel is beginning to turn a corner in its critical Data Center business segment. There is also optimism around its FPGA and dedicated AI silicon products such as Gaudi that are coming along nicely. New products will hit the market this year while others are taping-in.
Intel expects to deliver the next generation of Xeon processors, Emerald Rapids, later this year, in Q4. Emerald Rapids chips are already sampling to customers, and volume validation is underway. These chips will improve overall performance-per-watt efficiency and operate in the same power envelope as Sapphire Rapids.
Intel is benefiting from the well-tuned Intel 7 process with the Sapphire Rapids CPUs, and major customers are adopting the Sapphire Rapids CPUs.
Intel’s latest updates on their Data Center and AI products
Intel has just announced its second-generation E-core Xeon processor, Clearwater Forest, which will succeed Sierra Forest and is scheduled to be delivered in 2025. This chip will be the first Xeon produced on Intel’s 18A process or its second-generation RibbonFET process, which is expected to re-attain process leadership for Intel.
Intel’s P-core and E-core Xeons are remaining distinct product lines, and the company is committed to the Ecore lineup for the long run.
As for AI acceleration ASICs, the company’s Habana Labs division recently announced their next-generation Gaudi3 deep learning accelerator, a process shrink of Gaudi2. Intel projects that the total addressable market for AI accelerators will be at least a $40 billion market opportunity by 2027.
Intel announces updates on new Xeon chips and processes
Intel also announced its upcoming Xeon processors. The Granite Rapids, Intel’s next-gen p-core based product, is set to be shipped in 2024. It will feature MCR DIMM support for increased memory bandwidth.
Sierra Forest is Intel’s first-gen efficiency Xeon that will come with 144 cores, offering better core density compared to AMD’s competing 128-core EPYC Bergamo chips.
At this investor event, Intel made it clear that it believes CPUs rather than GPUs will dominate the total addressable market for AI silicon by 2027.
Intel is projecting a positive outlook as it continues to innovate in both its CPU and AI fields. It is apparent that the company is making significant strides in its Data Center and AI group, which is critical to its overall success.
Image Source: Wikimedia Commons
