feature article
Subscribe Now

RISC-V Fast Tracks Simpler Extensions

More Hope for Erstwhile CPU Designers

One of the charming aspects of RISC-V is that it’s so… flexible. As an open-source processor specification, absolutely anyone can use it, modify it, and commercialize it. There are no licensing fees, no rules, and no compatibility test. It’s the Wild West of processors. Up to a point. 

If you want to call your processor RISC-V, however, you have to follow the rules. The name “RISC-V” is trademarked, and there are also trademarked variations, subsets, profiles, and extensions that are all managed by the Switzerland-based RISC-V International organization. So, you’re welcome to ride off roughshod and blaze your own trail, pardner, but if you want to mingle with civilized company, you’ll have to tidy up and follow the rules. 

One of those rules governs how you propose new additions to the baseline CPU architecture. Again, you’re free to create your own add-ons any dadgum way you like, but if you want them to become part of the official RISC-V canon, there are procedures to follow and papers to fill out. It’s a lengthy process, but one that ensures that all officially approved extensions meet certain criteria and play nice with all the other extensions. 

That too onerous for you? Well, maybe there’s a middle ground. Starting today, if your proposed extension is simple enough, minor enough, and well-behaved enough, you might get special dispensation to “fast track” its approval. You’ll still have to dot some i’s and cross some t’s, but it’ll all be over much sooner and you can be on your way. 

According to Mark Himelstein, CTO at RISC-V International, the new fast track program includes all the same steps as the usual path – he hesitates to call it the slow track – but they just go faster. The idea is to allow relatively simple and uncontroversial extensions to get checked out and checked off without taking up so much time. 

As an example, the proposed new ZiHintPause instruction isn’t nearly as complicated as a complete suite of floating-point operations or a package of AI-acceleration extensions. The latter two are currently plodding their way through the review and verification process. But ZiHintPause is just a single instruction with a comparatively trivial function. It tells the CPU that this might be a good time to enter low-power sleep mode. Whether the hardware actually does anything at all is left up to the chip implementer. It’s just a hint, not a command. Apart from assigning it an opcode and documenting its intended purpose, there wasn’t a lot to do. Hence, its quick approval.

Himelstein says a fast-tracked proposal should take somewhere around two months or so to get approved, give or take a few weeks. There’s still the 45-day public review process, where interested parties can provide feedback (or not) if they foresee a problem. “That doesn’t scrunch down,” he says. But there’s less documentation required and lower hurdles for proof of concept. 

The overall procedure doesn’t change, either. All RISC-V extensions must be proposed by two or more dues-paying RISC-V members, who then form a committee to “own” that extension and shepherd it through the approval process. This minor hurdle prevents random citizens from spamming headquarters with every new idea. (They’re still free to implement extensions on their own, without any form of approval.) The relevant RISC-V committees then meet, discuss, revise, and approve the proposal, often without ever disturbing upper management. “We trust our committees,” says Himelstein. “We try to push power down. If we centralize everything, it slows down.” 

The ZiHintPause instruction was approved, by the way, making it the first official RISC-V extension to make it through the fast track process. Himelstein says there are at least a dozen more in the pipeline. So, look for RISC-V to keep changing. Just faster.

Leave a Reply

featured blogs
May 24, 2022
Today is going to be my monthly update. This normally runs on the last Friday of the month, but that's a Cadence Global Recharge Day, so we will all be off. For various other reasons, I need to... ...
May 20, 2022
I'm very happy with my new OMTech 40W CO2 laser engraver/cutter, but only because the folks from Makers Local 256 helped me get it up and running....
May 19, 2022
Learn about the AI chip design breakthroughs and case studies discussed at SNUG Silicon Valley 2022, including autonomous PPA optimization using DSO.ai. The post Key Highlights from SNUG 2022: AI Is Fast Forwarding Chip Design appeared first on From Silicon To Software....
May 12, 2022
By Shelly Stalnaker Every year, the editors of Elektronik in Germany compile a list of the most interesting and innovative… ...

featured video

EdgeQ Creates Big Connections with a Small Chip

Sponsored by Cadence Design Systems

Find out how EdgeQ delivered the world’s first 5G base station on a chip using Cadence’s logic simulation, digital implementation, timing and power signoff, synthesis, and physical verification signoff tools.

Click here for more information

featured paper

Introducing new dynamic features for exterior automotive lights with DLP® technology

Sponsored by Texas Instruments

Exterior lighting, primarily used to illuminate ground areas near the vehicle door, can now be transformed into a projection system used for both vehicle communication and unique styling features. A small lighting module that utilizes automotive-grade digital micromirror devices, such as the DLP2021-Q1 or DLP3021-Q1, can display an endless number of patterns in any color imaginable as well as communicate warnings and alerts to drivers and other vehicles.

Click to read more

featured chalk talk

Expanding SiliconMAX SLM to In-Field

Sponsored by Synopsys

In order to keep up with the rigorous pace of today’s electronic designs, we must have visibility into each step of our IC design lifecycle including debug, bring up and in-field operation. In this episode of Chalk Talk, Amelia Dalton chats with Steve Pateras from Synopsys about in-field infrastructure for silicon lifecycle management, the role that edge analytics play when it comes to in-field optimization, and how cloud analytics, runtime agents and SiliconMAX sensor analytics can provide you more information than ever before for the lifecycle of your IC design.

Click here for more information about SiliconMAX Silicon Lifecycle Management