Chatter around GPUs for RISC-V is growing

Power for the people, and AI

The activity around creating a legit graphics processor for RISC-V chip designs, an emerging competitor to x86 and ARM, is gaining steam.

Special interest groups at RISC-V next year will expand the focus on extensions for shaders and advanced matrix operations, which is important for artificial intelligence and machine learning, Mark Himelstein, chief technology officer at RISC-V, told The Register.

RISC-V International, which developed the instruction set architecture, has interest groups develop extensions that users can add to their chip designs.

In 2021, 16 RISC-V extensions were ratified, Himelstein said, and that number will grow next year. Many new extensions were part of mainstream computing chips announced this year at the RISC-V Summit.

"If somebody needs something more quickly, they can implement it as a non-standard extension and implement it themselves with their proprietary bits, and then bring it back to RISC-V, try to get it in as a standard, if that's what they want, or they could keep it proprietary," Himelstein said.

The RISC-V focus on GPUs is largely from an AI and scientific computing perspective, though the focus on shading represents a fledgling effort to create a full-fledged GPU extension for gaming and graphics design.

In the meanwhile other options for GPUs for RISC-V are available, though don't expect to buy them off the shelf.

Imagine all the GPUs

Imagination Technologies, which this week re-entered the CPU market with a RISC-V processor called Catapult, said its GPUs will work with RISC-V. The company's PowerVR graphics cores were used by Apple in iPhones, and are popular in mobile devices with chips from companies including MediaTek.

"There is no reason why you could not integrate C-series -- which is the part that has ray tracing -- with RISC-V," David Harold, chief marketing officer at Imagination, told The Register.

Andes Technology, which creates RISC-V chip designs, has verified that Imagination's GPUs work with RISC-V, and so has RIOS Lab, which has David Patterson, vice chair of the Board at RISC-V Foundation, on staff.

Imagination has a broader view of RISC-V and GPUs beyond PCs. There's a broader application alongside its AI, networking and other intellectual property in areas like automotive.

"We think that is the design future for the kinds of intense workloads particularly that are driven by things like autonomous systems," Harold said.

The need for a GPU on RISC-V could be fundamental as the chip architecture gains importance, Shreyas Derashri, vice president of compute at Imagination, told The Register.

"If you're running a face unlock algorithm or face detection, they are parts that run sequentially that can run better on a CPU with a vector extension. There are parts in floating point that will work better on a GPU," Derashri said.

There are other efforts to put a GPU atop RISC-V CPUs. A developer board with a RISC-V chip and a GPU went on sale, but sold out quickly. Another group is porting Nvidia's software development kit called CUDA, which is designed for the company's GPUs, to RISC-V.

Similar topics

Similar topics

Similar topics


Other stories you might like

Biting the hand that feeds IT © 1998–2022