Intel CTO needs builders to construct as soon as, run on any GPU

Over 20 years in the past, the Java programming language, initially developed by Solar Microsystems, supplied builders the promise of having the ability to construct an software as soon as after which have it run on any working system.

Greg Lavender, CTO of Intel, remembers the unique promise of Java higher than most, as he spent over a decade working at Solar. As a substitute of needing to construct functions for various {hardware} and working programs, the promise of Java was extra uniform and streamlined growth. 

The power to construct as soon as and run anyplace, nevertheless, is just not uniform throughout the computing panorama in 2022. It’s a state of affairs that Intel is seeking to assist change, at the very least in relation to accelerated computing and using GPUs.

The necessity for a uniform, Java-like language for GPUs

“Immediately within the accelerated computing and GPU world, you should use CUDA after which you possibly can solely run on an Nvidia GPU, or you possibly can go use AMD’s CUDA equal operating on an AMD GPU,” Lavender instructed VentureBeat. “You possibly can’t use CUDA to program an Intel GPU, so what do you employ?”

That’s the place Intel is contributing closely to the open-source SYCL specification (SYCL is pronounced like “sickle”) that goals to do for GPU and accelerated computing what Java did a long time in the past for software growth. Intel’s funding in SYCL is just not totally selfless and isn’t nearly supporting an open-source effort; it’s additionally about serving to to steer extra growth towards its just lately launched client and information middle GPUs.

SYCL is an strategy for information parallel programming within the C++ language and, based on Lavender, it appears quite a bit like CUDA.

Intel helps standardization for one code to rule all of them

To this point, SYCL growth has been managed by the Khronos Group, which is a multi-stakeholder group that’s serving to to construct out requirements for parallel computing, digital actuality and 3D graphics. On June 1, Intel acquired Scottish growth agency Codeplay Software program, which is likely one of the main contributors to the SYCL specification.

“We should always have an open programming language with extensions to C++ which might be being standardized, that may run on Intel, AMD and Nvidia GPUs with out altering your code,” Lavender stated.

Automated device for changing CUDA into SYCL

Lavender can also be a realist and he is aware of that there’s a lot of code already written particularly for CUDA. That’s why Intel builders constructed an open-source device known as SYCLomatic, which goals emigrate CUDA code into SYCL. Lavender claimed that SYCLomatic in the present day has protection for about 95% of all of the performance that’s current in CUDA. He famous that the 5% SYCLomatic doesn’t cowl are capabilities which might be particular to Nvidia {hardware}.

With SYCL, Lavender stated that there are code libraries that builders can use which might be machine unbiased. The way in which that works is code is written by a developer as soon as, after which SYCL can compile the code to work with no matter structure is required, be it for an Nvidia, AMD or Intel GPU.

Wanting ahead, Lavender stated that he’s hopeful that SYCL can change into a Linux Basis challenge, to additional allow participation and development of the open-source effort. Intel and Nvidia are each members of the Linux Basis supporting a number of efforts. Among the many initiatives the place Intel and Nvidia are each members in the present day is the Open Programmable Infrastructure (OPI) challenge, which is all about offering an open customary for infrastructure programming items (IPUs) and information processing items (DPUs).

“We should always have write as soon as, run all over the place for accelerated computing, after which let the market resolve which GPU they need to use, and stage the taking part in subject,” Lavender stated.