ARM’s Chinese Venture Goes Sour

We’re used by now to many of the more capable microcontrollers and systems-on-chip that we use having an ARM core at their heart. From its relatively humble beginings in a 1980s British home computer, the RISC processor architecture from Cambridge has transformed itself into the go-to power-sipping yet powerful core for manufacturers far and wide. This has been the result of astute business decisions over decades, with ARM’s transformation into a fabless vendor of cores as IP at its heart. Recent news suggests that perhaps the astuteness has been in short supply of late though, as it’s reported that ARM’s Chinese subsidiary has gone rogue and detatched from the mothership taking the IP with it.

It seems that the CEO of the Chinese company managed to retain legal power when sacked by the parent company over questionable ties with another of his ventures, and has thus been able to declare it independent of its now-former parent. It still has the ARM IP up to the moment of detatchment and claims to be developing its own new products, but it seems likely that it won’t receive any new ARM IP.

What will be the effect of this at our level? Perhaps we have already seen it, as more Chinese chips such as the cheaper STM32 clones are likely to get low-end ARM cores as a result. It seems likely that newer ARM IP will remain for now in more expensive non-Chinese chip families, but in the middle of a semiconductor shortage it’s likely that we wouldn’t notice anyway. Where it will have a lasting effect is in future Chinese joint ventures by non-Chinese chip companies. Seeing ARM’s then-owner Softbank getting their fingers burned in such a way is likely to provide a disincentive to other companies considering a similar course. Whether ARM will manage to resolve the impasse remains to be seen, but it can hardly be a help to the rocky progress of their Nvidia merger.

69 thoughts on “ARM’s Chinese Venture Goes Sour

    1. You DO know how far behind ARM RISC V is don’t you? RISC V is fine for microcontrollers, but that’s not where the leading edge IP is. It’s not going to get anywhere in the Application Processor space in the foreseeable future as you need a huge IP library for modern SoC designs (everything from Flash and DRAM controllers through high speed IO to GPUs). ARM has all this in a huge off-the-shelf catalog, with well understood interfaces for adding your own designs.

      RISC V isn’t open source in the same way as software is (a company is free to roll their own RISC V design, but then has absolutely no need to make that freely available) and this doesn’t mean that the catalog of IP will be available for all RISC V chips. Even low-end parts may be cheaper using the ARM platform because of this.

      1. “RISC V isn’t open source in the same way as software is (a company is free to roll their own RISC V design, but then has absolutely no need to make that freely available) ”

        I disagree with that, RISC V is open source in the same as much software is. Not everything is GPL’d, many open source licenses allow exactly the same “roll their own …. no need to make that freely available” for software.

        I agree on the delta with between ARM and RISC V, that feels like it will take a while to close.

        1. RISC V effectively is a published API. Non of the actual implementation is provided, only a spec and a set of tests to validate against.

          Open Source software means you have access to the source code, and can contribute or fork to your hearts content. If you can’t then it’s not open source.

          You get non of this with RISC V. You can’t see how company X has improved the ALU or company Y has tweaked the branch predictor. Because of this you don’t get the merging of ideas and components that makes OS software powerful. Look at what happens to an OS software package that has little contributor interest, it languishes or dies.

          1. “Open Source software means you have access to the source code, and can contribute or fork to your hearts content. If you can’t then it’s not open source.”

            But if you fork, not all open source requires you to keep the changes open – e.g. the BSD licenses. That’s the original poster’s point. I can see the code that Google Chrome is *based* on but I have no idea what changes they’ve made to it. Exactly the same.

            It’s open source and free, but not “free as in freedom” as the FSF defines it.

          2. I tried Debian in 2000, and gave it up.bevause it didn’t include Pine. Pine “wasn’t free enough” by Debian (and I guess GNU) standards. It was under some other license, the source code was available, but it clashed in philosophy (I can’t remember details).

            So I switched to Slackware, which did include Pine, and all was well.

      2. It already has an application board on the market, the HiFive unmatched. It’s really expensive (~$650) for a board somewhat less performant than the RPi4 (though it does have a cool realtime core folded in with the four application cores), but it’s out. There are also multiple linux-capable application boards from Pine64 and Sipeed among others that will come out within the year that will start at a claimed $12.50 (depending heavily on core count and RAM size).

        I’m not saying that Risc-V is at all in the same cost/performance regime as ARM for application processors yet, but the progress is shockingly fast.

  1. the acquisition of ARM by nvidia is such a charley foxtrot, such a clear indictment of western financial and regulatory structures, that the idea of china forking ARM seems like its last best hope for continuing relevance. but time will tell.

    1. Arm China is 49% Arm and 51% local investors, Arm never had full control of its chinese subsidiary. Maybe Softbank was too naive and did not learn from the mistakes of other companies as many similar ventures took such a turn in the past.

      1. Same happened to Google, They got into China only by starting research there, then their IP got stolen and they got kicked out of China again.

        And they did similar things to many companies. I find it quite surprising that such big companies fall for it.

        China is also a leader in “smart city” technology which is exported especially to countries with dictatorial regimes.

        China is very much abusing the psychology of people not caring as long as something does not hurt them personally, then doing something sneaky, and at some point you find you’ve been outmaneuvered.

        Some years ago China built a big trade center in the middle of Africa and gave it away for free, inclusive all computer equipment. A bit later the IT compartment discovered strange data exchanges between those computers and China. Gosh, who would have thought that possible. It was at about the same time as the “extra IC’s on the supermicro main boards.

        China is also buying up harbors all over the world. For example in Greece.
        Greece got almost kicked out of the EU after it leaked they lied about their financial situation, and which lead to a fall of about 20% of the EURO. After that Greece became easy pickings for the Chinese.

        I do not follow world news much and there are many, many more examples

        My only small role in this is buying cheap stuff from Ali which I still do.

    2. hmm, they set something up in china, gave it all the IP, then it went on its own way…
      Who on earth thought that there was even a possibility that could happen? They must have been deeply shocked and surprised. I mean, it has never happened before and China is very good at protecting overseas (to them) IP…

      Maybe I could sell those ARM guys a bridge I own..

  2. So, what you are saying, in essence, is that someone in China has stolen an IP and is now going to flood the market with cheap and, probably inferior product? Say it isn’t so! /S

    The last I looked, EVERYTHING China makes and/or sells, is based on someone else’s design or IP as they don’t come up with anything on their own. Most all of their technical prowess comes from stealing and copying someone else’s ideas or, technology. If anyone is surprised by this story, I would suggest getting out of the basement for a little bit and getting some sun before someone starts calling you “Powder”. :-)

  3. But the interesting thing is that it seems like most of the American innovation (and corresponding patents) is done by Chinese people these days. Does this imply that the Chinese people that want to innovate move to the US?

    I don’t know, so just a thought.

  4. the fact that china is avoiding a third world fate by learning from the western industrial example is not “unofficial”, it’s not “quiet” and there is no “plausible deniability.” they’re proud of what they’ve accomplished.

    if you look at the statistics of how many people were lifted out of poverty globally in the last 50 years, that isn’t a diffuse effect. the lion’s share of that fantastic result comes from the chinese communist party which is not shy about what it’s doing.

    1. The Chinese advantage is a populace that averages more that 5 IQ points above the western world. China would be doing even better if the government’s guiding principles weren’t theft and power. Bad government and bad culture can outweigh any natural advantage.

      1. While I’m bad at phrasing this exactly, I’ll try my best.

        Chinese education seems to be focusing hardcore on making “theoretical” scientists or other “book smart” type of people, and those who fails at that, get more “menial” jobs.

        Which ultimately leads to very few “real world experience” smart people in “smart” positions.

        Doesn’t help that their governing structure rewards obeying blindly and punishes those who asks questions, because inevitably they’ll ask the “wrong” ones.

      2. The American world maybe, which isn’t too hard to beat given the intentionally negative influence on education. It has been shown in many studies, that Asians in general, memorize and immitate more, due to cultural norms. They excell at standardized school tests, they are certainly not more creative, innovative, or more intelligent at all.

  5. “USA can’t do anything about it because China can ruin them and many other countries financially.”

    I imagine they would like to believe that, but China has a lot of it’s own issues to straighten out before they start swinging a stick they don’t have.

    1. China couldn’t develop a domestic Automatic Transmission until they acquired Australian Transmission Manufacturer DSI (Drivetrain Systems International) Formerly BTR, and even more formerly Borg Warner.

      The transmissions in question are the 4 Speed used previously in Ford Australia’s Falcon, and the Ssangyong Korando and Musso. As well as the newly developed (far too late to save the Australian business) 6 Speed used by Ssangyong.

      They took the entire Albury NSW factory and had it shipped to China. They even flew many staff to the new factory in China to teach the locals how to make the parts.

      The word coming back from China wasn’t good, and they were really out of their depth trying to manufacture such old 4 speed transmission designs let alone the 6 speed, and evolve the design going forward.

      I know because I worked at the Albury factory for 8 years.

    2. China has at least 1,1 trillion dollars worth of U.S. Treasury bonds. They have massive surplus in export against U.S. market and they control the value of yuan in relation to USD. China can easily use this to destabilize U.S. economy. China has its own problems, for example they have to follow 45th Rule of Acquisition: expand or die. Which they did in 2015 by devaluing yuan to boost their export. It was a small change but affected most markets. If China goes to economic war with USA, their economy will suffer. But it would be worse for the USA, and when their economy tanks, it will cause another global crisis…

  6. What the difference between what this Chinese company has done with ARM design contrasted with the fact Apple has a forever licence to do what it likes in ARM “improvements”?

  7. Genuine question here: does Chinese companies really lacks what it takes to research and develop new tech? Or there is no point in research something that will be manufactured there anyway?

    Not judging – I am really asking.

    See, when pandemics started, we had such a shortage of all sorts of things that were made there because “it was cheaper”. I’m pretty sure that some of you already had the surprise of sending an original PCB layout to manufacture, just to find it available later on those “alibababish” sites. Maybe local laws and/or international policies that creates some sort of IP blackhole?

    1. There’s a return on investment tradeoff that many companies make. Develop something new that’s difficult, then sell as many as you can before other people, having seen that you’re successful and reverse-engineered what you did, start selling stuff at a lower cost because they don’t have to fund anywhere nearly as much R&D, or be the other people reverse-engineering.
      There’s a decent ROI on only reverse engineering already successful products to undercut them, especially if your fab costs are lower.
      If you’re looking for a low risk, low but reliable return business, why would you develop new products?

      From this standpoint, IP laws are an attempt to stretch the period between your release and the release of the copies.

    2. “I’m pretty sure that some of you already had the surprise of sending an original PCB layout to manufacture, just to find it available later on those “alibababish” sites.”

      Ooh, now this is something I hadn’t considered. When I see Chinese copies of open source projects I just assume the downloaded the source and design files legally then made the copy. I hadn’t considered they might be stealing designs that weren’t shared as open source. Do you have any examples of this? I’ve only ever made PCBs at home, never ordered out for them, but there’s a few things I’ve considered using a fab house for… this could make me rethink that idea simply as a matter of principle. All of my designs go open source anyway.

          1. We had a project that runs a PCI bus off a FPGA. FPGA was not 3.3V tolerant, so we had to turn on the clamping diodes to 3.3V rail which was very light load. I look at the schematic and suggested replacing some of the decoupling caps on the BOM with resistors to add some loads.

            Fixing the schematic, get new 14+ layers PCB made, paperwork etc can take month+…

          2. I worked on a project where the consultant RF engineers didn’t respect the DC isolation of a mixer and used a resistor to couple to the next stage. It didn’t work.

            Then me (main job software engineer but degree in electronics too) comes along, reads the spec sheet, and replaced the R with a C. Board then worked. Red faces at design consultancy.

            At least the PCB didn’t need a redesign :-)

    3. The USA in particular (they had most to lose) loved getting cheap production in China, all they had to do was send over the details on what they wanted made, and most importantly, how to make it. It doesn’t take a genius to work out what happened next unless you are racist and totally underestimate the Chinese. I am not surprised the Chinese had other ideas that did not involve forever toiling in factories and saw the Capitalists as fair game.

    4. Do they lack what it takes? In general the answer is yes. For example the CCP’s big show piece state of the art semiconductor fab and initiative is bankrupt and the leaders under investigation for looting all the money. This plays out with every big project. And if the CCP has had any lasting effect on the culture, it is that you should never do anything that gets you noticed. That is not an environment that leads to innovation. The CCP took all Bob Ma’s equity in Alibaba and forced him to publicly join the party. This is not a good example for innovators.

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.