Qualcomm & Google – Starting low

Low volume, low value.

• Qualcomm and Google have announced that Android Wear OS will be supported on RISC-V processors opening the way for the first market implementation that takes RISC-V on Android beyond vapourware for the first time.
• Qualcomm is the dominant supplier of silicon for Wear OS but to date, this has only been available for Arm powered processors.
• This announcement extends support to RISC-V where Qualcomm will develop RISC-V-based silicon for Wear OS.
• At the same time, Google will ensure that its software will run optimally on RISC-V meaning that users should be indifferent to the processor architecture that is running their Wear OS wearable device.
• RISC-V has long been seen as a competitor to Arm and in certain market segments it is making headway at the very low end.
• The RISC-V instruction set is free from royalties but as usual in these sorts of arrangements, one gets what one pays for.
• This means that very little comes with a RISC-V licence meaning that the chip vendor will have to do a lot of work to produce a working processor design.
• The higher the performance requirement, the more work is required in order to create a chip capable of meeting the desired specification.
• This is where the economics of Arm come into play as buying a design from Arm for a high-end processor will often work out cheaper than doing all of the R&D required to get a RISC-V processor to the required level of performance.
• This is because Arm does the development once and then is able to amortise that R&D over a very large number of chips.
• This remains the case even including the healthy margins that the company earns.
• However, the lower the performance and functionality requirement, the more sense it makes to use RISC-V because the development cost is much lower.
• This is especially the case when volumes are likely to be high.
• This is why almost all of the traction that RISC-V has seen is at the low end of the market which is where it is likely to remain although it is slowly eating away at the bottom end of Arm’s position.
• Furthermore, the intention of Nvidia to acquire Arm from SoftBank galvanised interest in RISC-V which has remained strong even though the acquisition failed to close.
• This announcement by Qualcomm and Google is significant as the likelihood of RISC-V appearing in the digital consumer ecosystem has greatly increased and if it is successful, it may be considered for other devices where Arm currently dominates.
• This is a big if and the real question will be around 3^rd party developers.
• Just as Apple had to manage the transition from Intel to Arm with its M-series processors for 3^rd party developers, the same will be true for any transition from Arm to RISC-V.
• For Wear OS, this is less of an issue as the majority of functions are 1^st party, but if one wanted to create an Android smartphone using RISC-V, this would quickly become the dominant issue.
• This is why I think it is a huge leap of faith to assume that just because RISC-V works well for Android Wear, that it will work well for higher-specification devices.
• The net result is that this is unlikely to have any material impact on Arm in the short term although the competitive threat from RISC-V in low-end devices and China is growing.
• At the moment Arm looks like it is gaining value market share from x86 and proprietary processors faster than it is losing it to RISC-V as the gains are happening at the high end and the losses at the low end.
• This means that although it may lose share in terms of unit volumes, this is more than made up for in terms of price or royalty per unit.
• How well Wear OS on RISC-V fares will determine how far beyond low-end microprocessors it can go although I suspect that it may do a lot better in China than elsewhere.
• This is because the ideological struggle that exists between the USA and China is driving China to free itself from dependence on foreign technology and RISC-V is something China could embrace and make its own.
• This would cause technology fragmentation but with the state firmly in control of the economy, a standard implementation could be mandated forcing Chinese companies to follow.
• This has been on the cards for some time, but it has yet to materialise in a big way leaving Arm by far the dominant processor in the Chinese digital consumer ecosystem.
• Whether Google and Qualcomm can move the needle remains to be seen.