Koduri drives the 'gpu computing' revolution with intel 'ponte vecchio'

Amid the announcement of the Ponte Vecchio architecture, Intel's new GPUs for the HPC 'Exascale' market, Koduri appears to have been a key player in the 'GPU computing' phenomenon.

Raja Kodurii reflects on GPU Computing and the importance of Xe architecture in this direction

Intel's Raja Koduri helped kick-start the 'GPU Computing' movement 14 years ago, when he worked at AMD. This time, he is refining the strategy for a more heterogeneous computing world.

Fourteen years ago, Raja Koduri helped usher in the era of GPU Computing, a year before Nvidia's boss Jensen Huang spoke of the phenomenon. The idea was that GPUs can do more than just video game graphics, they can also solve scientific problems, which is starting to happen right now.

Nvidia has already implemented its technology for this purpose, especially for AI calculation and Intel seems to also want to enter this market with Ponte Vecchio and its Xe architecture.

Koduri commented in an interview to ZDNet:

That software movement, he said, is leading to a new era of computing that can be programmed to be as powerful as supercomputers.

Perhaps more important than the new hardware, is the provision of a beta version of its high-performance systems programming software toolkit, called OneAPI, which simplifies scheduling of supercomputing tasks across many types of processors. and systems.

When Koduri was at AMD, he used to say that Intel had a "buffet" lunch offer when everyone wanted burgers and a smoothie, which is the simple option that Nvidia provided.

Koduri declined to elaborate on the differences between the 'Ponte Vecchio' (Xe-based) GPU and other GPUs. All it says is "There are several modes of operation in this architecture that make it much more flexible than current GPUs with existing architectures." Koduri said, one can "map many more workloads on it, " adding, "We have a new way of doing vector processing in architecture."

Visit our guide on the best graphics cards on the market

"There are some details that we are not revealing at this time, " he continued. "The short answer is that their transistors are much more optimized" for HPC.

Koduri is confident that this new graphical architecture will be more flexible and perform better for AI or scientific computations that require large-scale processing, than what Nvidia is offering right now, for example. We will see to what extent it is true, but Koduri has been a visionary in this direction, where GPUs are no longer used simply for games, now they have a bigger role. We will keep you informed.

Zdnet font