Why do university projects often adopt RISC-V?

Author: Jim Turley Translated from EEJOURNAL

Sometimes it’s not about technology or performance

New microprocessor designs often appear in universities. The processor is replaced every few years. In the past, Liu Xingguo included RISC, VLIW, DSP and now RISC-V.

There are many reasons why RISC-V is popular. Why not ARM, or x86, or even MIPS, SPARC, 8051, Clipper, PowerPC, or PDP/11? RISC-V is almost the default processor of choice.

Because he does not belong to any commercial company.

This response came up several times over the past year as I interviewed chipmakers, software companies, researchers, venture capitalists, startups, and various members of the tech industry.

“Why did you choose RISC-V?”

“Because of our research funding.”

“Huh? Your financial backers say you have to use RISC-V? That’s very unusual.”

“No, because they said we can’t use a licensed CPU. By default, the only thing left is RISC-V.”

Some version of this conversation occurs almost every time a startup partners with a university, especially a European one. Universities are happy to work with industry players, who have funding, research and talent pools. But there is a catch, their involvement must be free of patents and licensing fees, and not proprietary technology.

The "proprietary" part rules out x86 processors, and the "licensing" ban eliminates ARM, MIPS, and other licensed CPU architectures. That leaves the field with RISC-V, the dominant open source processor design today. As a result, startups and universities are increasingly basing their work on RISC-V, as they have few legal options.

I believe that more ardent RISC-V enthusiasts will jump out and say that this is also because RISC-V is so flexible, scalable, powerful, modern, caring, good-looking, etc. But this is not the case, at least among the companies and groups I have talked to. They all agree that RISC-V is more than enough for their tasks, and they have nothing against it, but their choice is based on politics and laws, not technology, engineering or performance. They use RISC-V simply because they have to.

GreenWaves and its GAP8 and GAP9 processors are an example. The company collaborated with the University of Bologna in Italy. CEO Martin Croome said the chip startup received a lot of help from UniBo faculty and students, but the collaboration required a "non-proprietary architecture." This in turn drove GreenWaves' decision to develop GAP8 and GAP9 processors based on RISC-V.

The same is true for Tachyum, whose Prodigy processor can execute the instruction sets of several different processors through emulation, including x86, ARM and RISC-V. So why did they choose RISC-V in the end? Legality and licensing are obviously not factors, otherwise x86 and ARM would not be on the menu. But CEO Rado Danilak said that some of his potential customers are universities, "Academic contracts require the use of RISC-V and prohibit x86. So, we made a RISC-V converter."

These restrictions are buried in contracts and funding records, but there are several potential reasons. First, a government-funded agency doesn't want their research dollars spent on a locked-down project. They might be a hesitant commercial entity, and they might be afraid of running afoul of licensing or patent laws, especially since those laws vary from place to place and from product to product. Modifying or reverse engineering a microprocessor is legally tricky, even just for experimental purposes. If you end up discovering or developing something, so what? You want to make sure your work isn't hampered before you start, not after the fact.

A processor ecosystem rises and falls for a lot of random reasons, rarely because it’s the best technical solution for a given task. Sometimes good software support helps sell CPU chips. Sometimes a big customer anoints a supplier, and others follow. Sometimes it’s an unusual business model, or pricing strategy, or just bad luck. As with any form of evolution, unexpected external factors weed out others, leaving the survivors wondering what luck made them successful, no matter how excited they are. It doesn’t seem right to blame university guidelines or funding provisions in a donor’s will.

All in all, there are many factors that influence the choice of processor.