Following on changes made in June that moved Intel’s HPC unit out of the Data Platform Group and into the newly created Accelerated Computing Systems and Graphics (AXG) business unit, led by Raja Koduri, Intel is making further updates to the HPC group and announcing two new division heads.
The new organizational change within AXG splits the HPC business into two distinct groups: the Super Compute Group, led by Jeff McVeigh, and the Super Compute Platform Engineering Group, led by Brijesh Tripathi.
As general manager of the Super Compute Group, McVeigh is focused on CPU and GPU products and technologies that are specifically optimised for high-performance computing. Ponte Vecchio, the forthcoming Xe GPU that will be the centerpiece of the future Aurora exascale supercomputer, is an example. McVeigh has been with Intel for over 25 years. His LinkedIn profile still lists his position as vice president and general manager, data center XPU products and solutions.
As general manager of the Super Compute Platform Engineering Group, Tripathi is focused on the architectures, the design and the validation of the platforms that will be deployed in various HPC and supercomputing systems, including Aurora. Previous to this new role, Tripathi was vice president and chief technology officer with Intel’s Client Computing Group.
Bifurcating the HPC unit into two camps creates a product and business side that’s headed by McVeigh, and an engineering architecture platform side, led by Tripathi. “It’s very typical of how we architect and how we build our teams – on the datacenter side and the client side in general, the business focus and then the architecture engineering focus,” an Intel representative told HPCwire today.
Intel, of course, is gearing up to deliver one of the United States’ first exascale supercomputers, Aurora, at Argonne National Laboratory next year, in partnership with HPE. Aurora will leverage Intel Sapphire Rapids processors and Ponte Vecchio GPUs inside HPE Cray EX cabinets. McVeigh, Tripathi and Koduri are all involved in that effort as well as other exascale-type initiatives that have yet to be disclosed, the Intel spokesperson said.
The new “Super Compute” group names are certainly on-brand for the exascale era, although the name change calls up the fact that the HPC business has always played a delicate marketing game, where being too leading-edge (or leading-edge sounding) can be alienating to would-be customers. Over the last three decades, branding efforts across the segment have transitioned to using the somewhat more accessible-sounding term high-performance computing. HPCwire was born out of a publication called Supercomputing Review, in fact. But maybe times are changing. With AI (“AI”) everywhere, HPC and supercomputing are not as scary-sounding.
Prior to this latest restructuring, the HPC team within AXG was led by HPC veteran Trish Damkroger, whose official title was vice president and general manager of Intel’s High Performance Computing Group. The company confirmed Damkroger is still at Intel but has not yet disclosed what her new role will be.
In June when the HPC group was moved into AXG, the Data Platform Group was restructured into two new business units, the Datacenter and AI Group and the Network and Edge Group. Sandra Rivera is executive vice president and general manager of the former and Nick McKeown is senior vice president and general manager of the latter. Rivera, McKeown, and Koduri report directly to Intel CEO Pat Gelsinger, as does Greg Lavender, senior vice president, chief technology officer and general manager of Intel’s Software and Advanced Technology Group.
The intention of the HPC group reorg is “to make Intel and the HPC business run more harmoniously, and to get products designed and developed, and manufactured much more quickly and into the hands of our customers,” said the Intel representative.
This is a developing story.