Raja Koduri, Intel’s senior vice president and general manager of Intel’s Accelerated Computing Systems and Graphics Group (AXG), revealed on Twitter that they do not intend to use Xe-HP
Commercialization of the GPU lineup.Intel’s plan is to stop the company’s subsequent development of its Xe-HP series of server GPUs and will not bring them to the market. The reason for this seems to be that Intel has evolved its original plan for the Xe-HP series into the HPC series, also known as Ponte
Vecchio, and Xe-HPG series, also known as Intel Arc.
This means that adding a set of GPU products to the original Xe-HP architecture makes the old plan meaningless.IntelThe flagship server GPU series is code-named Arctic Sound. With the re-emergence of Intel’s HPC department, the progress has been “under development”.
Koduri earlier demonstrated the chip’s true capabilities in Intel Labs. As the first high-performance Xe chip under development, semiconductor giant TSMC has already manufactured bot Xe-HPC and -HPG molds for Intel.
Due to the lack of reports on Intel’s Xe-HP series, the outside expects that the company’s internal focus of its discrete graphics business will change. Intel previously demonstrated the Xe-HP chip, which can provide up to 42 TFLOPS of FP32 throughput, while providing samples to a small number of customers. But in the following year, Intel shifted its focus to R&D and production of Ponte Vecchio for their Aurora supercomputer customer base. The company also plans to release the Xe-HPG Alchemist GPU series in the first quarter of next year.
Koduri said that the Xe-HP series has been fully utilized and continues to be used as a “development tool for Aurora and Intel’s oneAPI system.”
The biggest question at present is that since this refocusing of R&D focus basically overthrows the enterprise-class GPU product line, will Intel be able to maintain their relationship with competitors NVIDIA andAMDThe competitiveness between the two is worthy of attention.