Intel Drops Xe-HP Server GPU Plans, to Stick with HPC and Client Graphics
Intel has dropped options to construct Xe-HP server GPUs commercially. This line of products and solutions would have powered cloud-based graphics rendering instances, for cloud-gaming or cloud-rendering apps. An announcement to this outcome arrived from Raja Koduri, overseeing the growth and monetization of Xe. Koduri mentioned that Xe-HP centered circumstances had been at first established up to electricity Intel’s oneAPI devcloud as a application improvement auto for oneAPI and the impending Aurora supercomputer of the Argonne Nationwide Laboratory.
The organization will now aim on Xe as a compute accelerator in the variety of Xe-HPC “Ponte Vecchio,” and discrete graphics in the customer phase, leveraging the Xe-HPG graphics architecture. The smallest derivatives, the Xe-LP, powers integrated graphics options observed in the company’s Core processors (11th Gen and afterwards). Back again in the August 2021 Architecture Day presentation, Intel’s technical quick for Xe HPC uncovered that the silicon by itself capabilities sure on-die hardware pertinent to graphics rendering (additional here). This would have gone on to energy the Xe-HP server GPU methods.