Last year, Jeff McVeigh, Intel’s readily available leader of the high-performance computing group, suddenly went silent, with no interviews granted or appearances at press conferences.
It led to questions — what’s McVeigh up to? Is he still at Intel? After all, he went silent after Intel made wholesale changes in its HPC business, including converting Falcon Shores to a GPU-only product and axing Rialto Bridge.
There’s an answer — McVeigh is now corporate vice president and general manager at Intel’s Software Engineering Group, according to his LinkedIn profile.
McVeigh previously was vice president and general manager of the Super Compute Group. He also featured prominently in Intel’s late 2022 decision to split the consumer and enterprise portions of the Accelerated Computing Systems and Graphics Group.
Jeff McVeigh was appointed interim leader of the AXG group, which merged into DCAI (Datacenter and AI Group). The consumer section of AXG was merged into a client group, and division leader Raja Koduri was reassigned.
In fact, the Super Compute Group — which handles HPC — hasn’t appeared in any meaningful Intel material since the International Supercomputing Conference last year.
It’s also very possible that McVeigh’s move to software means Intel may have dismantled its HPC-focused Super Compute unit. McVeigh took on his new role in Nov. 2023, just around the time Aurora finally huffed and puffed its way to the Top500 list for the first time.
The division’s main product – Ponte Vecchio – made its way to Aurora but has failed to find adoption elsewhere. Aurora took eight years to get up and running, partially due to Intel’s inability to deliver Ponte Vecchio on time.
Sadly, Ponte Vecchio’s significance has devolved into a cheap talking point for Intel executives looking to score brownie points at events.
“It’s the Super Bowl of integrated design,” said Stuart Pann, who runs Intel Foundry Services, during the Intel Foundry Direct Connect 2024 event.
The DCAI has further realigned in recent months.
Earlier this year, Justin Hotard was appointed executive vice president and general manager of the DCAI group. Hotard, formerly of HPE, replaced Sandra Rivera, now CEO of Altera, Intel’s FPGA spinoff.
Intel has specifically been tapping into former HP executives to run its most important divisions. Stuart Pann previously managed HP’s supply chain, and Intel has also hired former HP comms leader Karen Kahn to lead the communications and PR group.
The chip maker has preferred to promote internally, but the outside appointments hint at Intel CEO Pat Gelsinger changing company culture as it moves into a foundry-first business model. Intel’s foundry business requires political and supply chain savvy as it is intertwined with national security issues surrounding semiconductors and AI.
While Intel is meeting its foundry goals, its chip business is still floundering.
Intel’s PC chip Meteor Lake, released last year, opened to bad reviews, with critics bashing its battery life and performance compared to Apple’s ARM-based and AMD’s x86 chips. Meteor Lake is based on chiplet technology but has flaws. The underlying message: customers should wait for a second-gen chiplet processor to iron out kinks in Meteor Lake.
On the server side, Intel’s Emerald Rapids chips, released late last year, are serving as a bridge to Granite Rapids, which is based on a new architecture and process.
Intel was slow to read the GPU and AI market and is now reshaping its GPU roadmap. Its next major GPU release, Falcon Shores, is planned for 2025. The chip maker is also prioritizing its Gaudi AI chips, which are being merged into its Falcon Shore GPU.
However, with its recent Blackwell GPU, Nvidia has taken a solid lead over Intel in AI. Nvidia is prioritizing its GPUs for AI and mixed-precision computing, and Intel is following in its footsteps.