Arm and NVIDIA – The unthinkable pt. II.

RISC-V has a second lease of life.

• While there is no sign that NVIDIA will be unable to maintain Arm’s crucial independence, just the mere fact of the possibility has breathed new life into competitors that prior to this deal looked pretty moribund.
• Arm’s biggest competitor is not Intel but RISC-V which is an open-source version of what Arm offers.
• RISC-V is often portrayed as a viable competitor to Arm but the reality is that when you don’t pay for something you end up with a lot less.
• This means that one has to do a lot more work to create a working chip using RISC-V meaning that a straight comparison of the royalty that one pays for the IP is not a good way to judge the cost-benefit of each offering.
• This raises the likelihood of fragmentation where chip makers fill in the missing bits with their own technology resulting in a series of incompatible chips, despite being based on the same processor.
• Furthermore, the RISC-V ecosystem is a tiny fraction of the size of Arm’s meaning that its products and services will tend to be more scarce and more expensive compared to Arm.
• That is not to say that the RISC-V ecosystem cannot be built but at the moment it is at a large scale and brand disadvantage.
• In 2019, there was a huge hype cycle for RISC-V driven largely by the increasing tension between the West and China but this started to flag towards the end of 2019 and in Q1 2020, some of its biggest proponents looked like they were on their last legs.
• RISC-V’s best-known proponent is SiFive which is effectively a design house that clients can go to design chips that are based on the RISC-V architecture.
• Towards the end of 2019, SiFive was rumoured to be on its last legs but managed to raise $65.4m in June 2020 from big backers including Qualcomm ventures.
• This round was announced in June which significantly predates the NVIDIA acquisition meaning that the sudden increase in interest was probably more about China than it was Arm’s independence.
• SiFive also managed to attract a senior Qualcomm executive to be its CEO in September in another sign of its rise from the grave.
• With the current trade and technology tensions between the West and China, China’s drive for technological independence is gathering pace meaning that it could seek to design out both the x86 and the Arm processor from its domestic industry.
• The biggest obstacle to doing this is the legacy software that already has been written which would then need to all be re-written.
• This is where China has an edge over other regions thanks to the influence that the Chinese Communist Party (CCP) has over the economy.
• If the CCP decides that the only way to secure its technological independence is to abandon Arm and go with RISC-V, it will be in a position to compel companies to shift their hardware and software.
• I think that the CCP is very far from doing this as, despite its control, this is a very risky move that could jeopardise the viability of a lot of Chinese companies.
• One only has to remember the disastrous TD-SCDMA Chinese proprietary standard for 3G networks to understand that the CCP will think long and hard before making such a profound move.
• However, I think it is this possibility that SiFive is betting on although there are also some opportunities for RISC-V in 8-bit controllers at the lower end of the market outside of China.
• A resurgent RISC-V is the last thing that Arm needs at the moment because the very messy situation in China with its joint venture Arm China could derail the entire acquisition as Chinese regulatory approval is required.
• This is one of the stranger disputes that I have seen in 20 years as a technology observer, but I think that at the end of the day, it boils down to a disagreement about money.
• Hence, regardless of the merits of the dispute between Arm and the sort-of ousted CEO of Arm China, I suspect that Arm will end up paying him a settlement to make this problem go away.
• Hence, I don’t think this will cause a long-term problem for the acquisition and Arm and NVIDIA very wisely set an 18-month time horizon to get the deal done.
• The end result is that this deal has undoubtedly given a shot in the arm to RISC-V.
• It is not surprising to see one of Arm’s biggest customers and supporters (Qualcomm) hedge its bets in this way.
• I still don’t think that RISC-V will amount to very much, but should China decide that RISC-V is the way that it is going, then this will change the whole ball game.
• This is going to take a long time to play out.