Why RISC-V must get its messaging right on open standard vs open source The possibility of America placing sanctions on RISC-V has increased the pressure on RV's governing body and its partners to get their messaging right about what this technology really is. One of the primary tools of the US and China in their trade war is placing sanctions against each other. The United States and its allies …
The point is that when talking to politicians, you must use words they and their voters understand, preferably in soundbites. By now, many people who care about computers and ICT know about "Open Source", but they do not understand "standards" and "ISA" even less. So, whatever you say, the politicians will hear "source" not standards.
> the ISA bus came out with the IBM PC and hasn't it been obsolete for years?
There was a LPC port (ref: Low Pin Count Interface Specification) used on some PC boards, typically LPC was used for communications with a plugin TPM (Trusted Platform Module)! LPC from a software standpoint looks like ISA! The pin out it different, but at it's heart is it ISA (only serial, instead of parallel)!
No, very definitely RISCV has been hoist by its own petard. What they’ve sold it as, is open source, not open standard. Open standard means “openly accessible and free to use by anyone”, it does not mean “openly extensible”. most “open standard” will tell you that being openly extensible is not a standard at all. SDRAM, USB, .png are open standards. DDR4 SDRAM with some extra pins to do on-memory computation….is not DDR4. A .png file with some extra codes that allow special things like grid warping….is not a standard .png file.
The RISCV people decided to pander to the open-source factions, and add “openly extensible”, which intrinsically fragments the ecosystem and makes it not-a-standard-at-all. That side of it is nothing at all to do with open standard, but it’s the major part of their offering to the public. Yes I perfectly understand that it isn’t open-source because there’s no source, and no requirement that the RISCV cores shall be open-source. As do the regulatory authorities.
It’s a brand new thing, openly extensible, which nobody else does because it was obvious to everyone else but the enthusiasts it was f*ing stupid thing to do. And if you ask how “openly extensible” shall be treated, the answer is “exactly like open source”. Because, it is a Intellectual Property on which one can build more stuff. It is a partially-processed raw material. So, the regulatory authorities have made exactly the correct ruling on that. Now, how *open source* should be treated, I neither know nor care.
Or some politician decides the same applies to Open Source software and bans "export" of Linux to China or blocks Github for users from countries they don't like.
It's the 1990's cryptography debate all over again. That ended up with the decision that cryptography code (as opposed to devices) was protected by the US' 1st Amendment. The RISC-V spec will undoubtedly end up the same, after fat headed and ill willed politicians have pissed in everyone's soup for many years.
I suspect the U.S. government could make life pretty miserable for anyone involved with the RISC-V Foundation if they wanted to. They could be put on trial for collusion or simply be denied visa entry to the U.S.
Since many RISC-V users and members are volunteers I doubt many of them will find it worth the risk.
The big winner in all this could be ARM, the UK company spun-off from Acorn Computer. I believe they're already quietly corresponding this to their prospective clients.
There's certainly more tension in the world, and less need for Europe to remain in America's orbit. However, both Europe and her colonies are Western civilization, and we will continue to be natural allies. And I really don't see America "sinking" more than anyone else is. For example, the PRC projects a strong image, like the USSR did, but I have no doubt it has the same issues beneath the surface.
They don't like it because they can't control it and make money from it in the American way.
So they are causing polarisation, undoubtedly to the relief (intended or not*) of those invested in Arm's fortunes.
RISC-V is forecast to continue to increase its market share, with or without the US marketplace. The actions of US politicians is ultimately detrimental to their electronics industry.
* It seems suspicious to me that this comes hot on the heels of Arm's flotation, but I don't have any knowledge, and I'm not a conspiracy theorist.
" It hopes to avoid a crackdown by getting lawmakers, policy wonks, and officials to understand what the community sees as the subtle difference between open source and an open specification."
Good luck with that. The people they're trying to persuade are either identical wit or from the same batch as those who think an encryption scheme can be back-doored so the good guys and only the good guys can intercept it. Some of them may think that mathematicians have made an error by not having π equal to 3.
"RISC-V must get its messaging right on open standard vs open source"
This heeadline makes me cross. RISC-V International have been clear for as long as I've been following RISC-V's progress (at least 6 years) that RISC-V is a standard, and that as such, companies are free to use it to produce processors, and that it is up to the company whether they open source the chip design or not.
Nonsense. To show that RISCV (and you) are just inventing distinctions that don’t exist, to hide a conclusion you want to reach, is trivial.
It’s a matter of recorded fact that open *source* RISCV CPU projects do exist, ie the Verilog is available on GitHub. Then, are you OK with all the measures you complain about, instead being targeted against *all open source RISCV CPU projects* rather than the open standard? Probably not, right. That in itself tells us a lot, but it’s still not my point.
Then, if regulatory did decide that the RISCV open standard project was fine, unencumbered, but the open source CPUs had to stop……is that a standard which you would have any remaining interest in? Which you would bother supporting or commenting about? To be clear, this would be a world in anyone could release and sell a RISCV silicon chip, or even a RISCV IP core as along as it was unmodifiable and you couldn’t see the source VHDL. But no sharing source code and academia/hobbyists couldn’t “work on RISCV” unless they individually built it from scratch and didn’t share the source. Is that project, as described legally prescribed, anything you could support or be interested in?
Be honest, no it isn’t. You want open source. RISCV are selling the whole project as open source able. Stop this shit pretence about it being open standard, and the Manglement don’t know the difference. Everyone knows the difference, and nobody has interest in RISCV as open *standard* without *source*.
it's fair enough to ask the questions, but not to assume you know the answers too, and then comment on your assumptions.
The reason we know that RISC-V is an open specification, is that it's published and freely available for anyone and everyone to inspect and use.
It's also widely known that the ISA was created in order that students could do some practical work in developing processor designs, without infringing anyone else's Intellectual Property. And for that point alone, RISC-V serves its purpose, and contributes hugely.
You generously point out that there are a number of open source RISC-V processors already available, from companies like Western Digital, OpenHW and T-Head. There are also already millions of proprietary RISC-V processors in tech since 2017, when nvidia started using the ISA, 2019, when Qualcomm started using it, and 2020, when Samsung followed suit. It's suggested that there's RISC-V in Apple products too, but I haven't the evidence to support that.
So yes - there is a very real distinction to be made between open source RISC-V processors (which will continue to arrive, at a slower pace), and their closed source counterparts.
Quite frankly, I don't really care what the US tries to do about RISC-V. As I've said any number of times on this forum, any action the politicians take only does its own industry injury. RISC-V is a thing, it's prodicted to grow its market share because it can beat ARM on PPA. US tech companies not only want it, but know they need it to remain competitive.
The bottom line is that the US government wants RISC-V too - they just don't want anyone else to have it, because it levels the playing field.
But the US can't stop RISC-V's use any more than they can prevent people from using alternatives to Microsoft or Apple software.
And the reason that RISC-V is developing so well, so quickly, is that people, educational institutions and businesses are choosing to collaborate to improve it.
But you are right about one assumption: I would prefer an open source processor over a proprietary one.
U.S. policymakers have (IMHO correctly) identified RISC-V as an important technology and standard which will allow China (and Russia and other "Axis of Evil" powers) to keep up with and make use of U.S. technology developed for this standard (operating systems, compilers, software).
If the Western world were to transition entirely to RISC-V (which isn't too far fetched) the entire technology stack would be based on this standard. All software, including sensitive dual-use software, would run on this platform and sanctions couldn't effectively be enforced since they're able to recreate hardware based on this standard.
Compare this with the x86 standard today. China isn't able to clone this standard and export hardware based on it to our shores because it's licensed technology. If RISC-V became the "new x86" China would be able to side-step all sanctions and at the very least continue exporting to our markets or (if we banned them) to other markets where the technology would still be wanted since it would run all Western software.
RISC-V will be like x86 then, a collection of CPUs which are incompatible beyond a common base instruction set. Have a look at the absolute train wreck that SIMD is on x86. There are numerous incompatible extensions, with Intel being the worst offender as they attempt to segment the market in order to extract the maximum revenue from each perceived category of users. It's marketing department driven design rather than technology limitations.
RISC-V will be the same. There will be a core instruction set with vendor specific extensions. Nobody will use the vendor specific extensions except for people who have a specific use case.
Arm does the same using proprietary function units. This doesn't stop most software from running on Arm chips though, as they just stick to the core instruction set.
> Compare this with the x86 standard today. China isn't able to clone this standard and export hardware based on it to our shores
I fail to see where the export part is important in any way in the context of sanctions and tariffs. If China thinks it needs to run x86 software, and can't get x86 chips due to sanctions, do you really think they won't make their own x86 compatible chips anyway, regardless of what the West says?
Also look up VIA for a bit of history around exactly this. The instruction set isn't the important part, the actual technology inside the chip is what is truly important for performance.
In an adversarial situation the notion of "not being able to do 'X' because its not licensed' becomes moot.
The reason why x86 won't get cloned -- which it already has been, BTW -- is that its not worth the effort. The instruction set is a bit of a dog's breakfast of adaptations and bolt-ons dating back to an early 8 bit processor. Its widely used because its there and highly developed but its definitely well past its prime.
"Compare this with the x86 standard today. China isn't able to clone this standard and export hardware based on it to our shores because it's licensed technology."
Actually, yes they can, because Shanghai Zhaoxin Semiconductor Co. bought those rights from Cyrix/Via. Their latest flagship is on par with a Kaby Lake i5. It is not going to win any speed records, but it would certainly be very usable for a lot of use cases.
They are legal to buy. They are readily available on sites like Aliexpress and Amazon. People don't buy them because it involves spending more money to use more electricity to get the same performance. Usually, low-end things either go for cheapness or power efficiency, so to buy Zhaoxin and get neither is not a popular decision. You can, though.
You might just as well sanction the use of 'English' – it is a 'standard', and it is 'code'.
There's nothing to stop the great US legislature banning the use of 'English' in dealings with 'foreign' interests on the grounds that 'important technological information may be passed on'.
Self-serving stupidity is shameless and incorrigible.
China has their own "standard" ISA called Loongarch, which is basically a fork of an open version of the MIPS64 ISA (during the short time when it was an open source ISA) combined with a few ideas from RISC-V and some custom bits added on since and still evolving.
If there's any pressure applied on RISC-V all that's going to do is get China to officially embrace Loongarch as their national standard which would probably help them, versus having efforts split between that, RISC-V, and ARM.
Encryption was once the province of proprietary and often ad-hoc algorithms. These eventually shook out to a relative handful of standard techniques which the American DES was most widely used. Unfortunately the US placed restrictions on the export of equipment that included encryption (you needed a license to export a WiFi access point, for example) so when the industry came to adopt a new global standard it made sure that any standardized technology was out of the reach, and so control, of the US government -- in fact, of any government.
The same concept is used elsewhere, including with RISC-V. The US government desperately wants to control it but since its an idea rather than a tangible 'thing' there's nothing to control.
(Ironically, one of the foundational strengths of the US itself is the notion that its Constitution is based on an idea rather than evolving from 'the divine right of kings'.)
"Ironically, one of the foundational strengths of the US itself is the notion that its Constitution is based on an idea rather than evolving from 'the divine right of kings'"
This is true. But to add irony of their own, too many citizens have elevated The Constitution to a position of majesty in its own right. For document with origins in the 18th century it is a remarkable achievement that merits a great deal of praise. But it should not be used as a gpld standard for evaluation of every new idea in politics and governance. It may be that XYZ is judged to be incompatible with the constitution, but that doesn't make XYZ morally wrong per se.
It can be changed back, but their points usually include the following:
1. You can't just pass a law, because it will be invalid. If you want to change something, you have to pass an amendment first.
2. It is hard to pass an amendment, so you need to plan for how you're going to accomplish that.
When people argue on that basis, they aren't saying that it is impossible to change it, but that you cannot act as if that change has already occurred.
The fact that someone would make this statement, and seven people would agree should scare you. William Shakespeare wrote plays in the 1500's the concepts of which still are still relevant today. Sam Clemens aka Mark Twain wrote books articles and such that are as relevant today as they were in the 19th century.
The US constitution is the gold standard, at least the words should be. It defines limits to government, and simple to understand rights. Dismissing it because it "was written in the 18th century" is like dismissing the evil 10 commandments because they were written before Christ was born. I keep forgetting how concepts like "Thou shall not kill" and "Thou shall not steal" are antiquated and should be ignored because everything is relative these days.
People better wake the f*ck up or these dystopian futures of Sci-Fi hollywood are going to be a reality. There are groups of people pushing society to a breaking point so they can "fix" it and take over every aspect of your life. I'll be dead by the time it matters but you young people out there should spend a little more time reading some of the documents created in the 18th century and earlier, before before you talk your way into an enslavement you were not expecting.
Risc-V is the tip of the iceberg. Groups of people that are in, and around governments around the world are pushing very hard to control everything you, the unwashed masses, want to do.
The statement you're disagreeing with and the statement they made don't appear to be the same statement. They didn't say that it was irrelevant because it was old. They said it wasn't perfect, and one of the reasons was that it was old. It wasn't perfect at the time, which many contemporaneous documents will demonstrate, so we shouldn't expect it to be perfect now.
All your other examples are subject to the same critiques. None of them were perfect, and many of them are less relevant now than they were at the time, if only because the things they talked about are different, and their discussion of them only considers the past situation. That doesn't make any of them worthless. Putting a document on a pedestal will not help because you need to rigorously consider which parts are still relevant and which ones could do with an update, and neither dismissing it as obsolete nor extolling it as perfect will help do that.
Not noted here is that the fastest RISC-V general purpose machines you can currently buy use the THead C910 core, which is:
1) Chinese
2) actually Open Source (except the vector unit): https://github.com/T-head-Semi/openc910
The fastest off the shelf RISC-V machine currently is the Milk-V Pioneer using the SG2042 SoC which has 64 C910 OoO cores running at 2.0 GHz, with 64 MB L3 cache and up to 128 GB RAM.
Of course this situation changes very fast. There will be several machines using SiFive's P550 cores in several months -- most from Chinese companies, or at least using Chinese SoC (SiFive's own HiFive Premier P550 board). And then at the end of the year the Milk-V "Oasis" (and others from at least Sipeed) using SiFive's P670 cores, but again in the Chinese SG2380 SoC.
There are a several US startups who started work on RISC-V core in 2021-2022 who will have much faster (Apple M1 class or better) cores, but those won't arrive in machines you can buy until 2025 or 2026.
XiangShan V3 (goal: 16.7 SPECint2k6/GHz, current simulations with RV64GCB: 14.7 SPECint2k6/GHz): High performance out-of-order core with RV64GCBV_Zbb_Zicond and 128-bit vectors, that is developed and validated by a Chinese university in collaboration with a bunch of industry-partners. The prior generates were taped out, and that's also the goal for this generation.
SiFive P670 (12 SPECint2k6/GHz, supposed to be comparable to the Cortex-A78): It's an out-of-order with RV64GCBV + vector crypto support and a 128-bit vector length with vector double-dispatch. It should be released with the SG2380 at the end of this year. The SG2380 has 16 P670 cores, as well as 4 SiFive X280 cores, that are used as an NPU. The X280s are in-order cores with 512-bit wide standard vector extension + custom in-vector register matmul extension, that can together supposedly achieve a total of 32 int8 TOPS.
China loves RISC-V and are investing in the future of RISC-V. What they are doing would be impossible in the west, it would be like ARM, AMD, NVIDIA, Intel, IBM and all the US technology universities all working together on the one CPU line. Is it crazy ?
The problem started with a ban on Huawei HiSilicon, to use ARM v9 set. Therefore all China tech goes on Risc-V implementation & development, almost 90% of development are from China, & at the moment, they have 11 of the 13 CEO's on the board.
US are tried to take Risc-V, thru Intel, to buy them and block China tech, like they did thru SoftBank on ARM. They not succeeded, and Risc-V are based in Switzerland now.
Surprisingly, RISC-V chips are known to be a generational 1.5 times ahead of existing ARM chips, when it comes to deployment requirements of AI algorithm models, which further fuels the adoption for it.
Do not forget, Huawei Ascend 910 Ai card was World's No1 in the 2019, and the ban on ARM & else is introduced in a same year.
"Chinese have been making huge leaps forward. They built their own spacestation, landed on the dark side of the Moon, (& on Mars), and started fabbing 7nm chips. If I was the US I'd be fucking terrified of what happens in the next 5 - 10 years"
Peshman
* Chinese Startup Unveils The First RISC-V Based AI CPU
https://wccftech.com/chinese-startup-unveils-first-risc-v-based-ai-cpu-powers-k1-domestic-laptop/
There are things I'm afraid of, but rapidly improving technology from a foreign source is not one of them. Fear like this is what starts wars. My fear is the generation of fear by the media and government about things like this will lead to physical responses. This will be bad. People are stupid, and want to tell other people how to live.
In retaliation, China has banned certain Western-made semiconductors from the country, and has been supporting efforts to make everything it needs in the Middle Kingdom itself.
China first has been going on for quite some time and it's aim is for everything China needs to be made in China.
PS Search Engines China first is not One China they are quite different!!
RISC-V has outfits in the US, China and Japan. Its upcoming conference is to be held in Germany, EU. Its officers are culled from the four corners of the globe.
The only thing the US has going for it is the registered office.
How hard can it be to register a new org in Dubai or somewhere, and reassign all the IP while Congress is still dithering?
The amount of words used to say nothing in the article is quite stunning. This is following the same pattern of so many other industries. It's a honeypot to attract useful idiots. Put out the lure, in RISC-V's case initially some high minded global community nonsense to appeal to those in place on the board at the time along with all the funding for this great act of humanitarianism we call RISC-V. In the process a second trap is set: the IT Bro's unending quest to be the first at the next big thing. This is of course the crowd that brought you monkey portraits and virtual real estate, so you really don't need to put in all that much effort. Before long, you have a legion of useful idiots doing all the initial setup work for you: everything from spurring the hype of the next big thing to even greater heights, contributing millions of lines of code, and getting the product placed all across the world. Then, as has happened time and time and time again: snap. The trap closes. Nothing on the upside of the Chinese maximum wage and state subsidies is able to compete with the hardware being shoved out in unending lines of e-waste. A national stated goal of cornering and reigning in this new market is no longer even a subdued goal as were past efforts to control the worlds surveillance camera market as an example. They will of course gain all the related industries and production in orbit around RISC-V chips, many probably being the tried, tested and proven 49-51 Joint Venture. Ultimately the people that made the initial move to Switzerland will mostly end up with nothing while a few with a good deal of money. The missionaries of the coming of the age of RISC-V will find themselves having been bug beggars to their own demise as the few RISC-V shops outside of China layoff and closeup shop. Furthermore overall tech jobs in the field of processors will have been shrunk and eroded by the new chips: x86 and ARM will have less and employ less. Don't forget, ARM China is still very much a two company one name situation. And AMD and Intel have both been so helpful on the x86 front. Along with that quiet 3rd holder of the x86 patent. This has all happened before.
"That's not to say all those in public life are too dumb to understand"
I think that's EXACTLY what this says. It doesn't matter if it's an open spec or open source; it's ludicrous to think a sanction can be placed on an open either one just describing an instruction set. These are a bunch of olds trying to regulate things they in no way understand.
In addition, whether it's open source or open spec, it was developed globally, not strictly within the US. If the US wants to pretend they can police the world; well, someone outside the US can simply fork the spec, have a spec that is identical except being emitted from outside the US, and freely give their identical but forked spec to anyone they want without consequence.
Regulating a concrete implementation (specific chips, or VHDL description of chips)? Silly but at least makes sense. Regulating an ISA? Completely ludicrous and pointless, particularly given it's widely available already, AND in fact already under production by mutliple Chinese companies.
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2025
