12-09-2022 | | By Robin Mitchell
Recent reports show that Intel has continued to show interest in RISC-V, and while many speculate that the support for RISC-V could help Intel gain better market share in mobile computing, it is also possible that RISC-V would allow Intel to also enter the microcontroller market. What challenges does Intel face against ARM and RISC-V, how could RISC-V help Intel enter the microcontroller industry, and would Intel ever leave behind their x86 instruction set architecture?
While numerous CPU architectures have been developed over the past 50 years, none have the market share and fame that the Intel x86 has boasted. This popularity was undoubtedly fuelled by the use of Intel processors in the IBM PC which was eventually cloned by other manufacturers and became the de-facto standard of the computing industry. As the Intel x86 ISA is proprietary, only competitors who pay Intel a substantial royalty can manufacture alternative processors, and this has led to only one competitor, AMD. However, the two companies now have many cross-license agreements thanks to AMDs development of x64 (which Intel relies upon), meaning that the two companies are in somewhat of a Mexican stand-off.
In the early 2000s, Intel primarily focused on microprocessor systems aimed at desktops and servers with little attention to mobile processing platforms. When mobile devices started to take off, designers were presented with a unique opportunity; they could choose any ISA they wanted. The reason for this opportunity arose from the fact that mobile devices do not need to run mainstream operating systems nor do they need to be compatible with the IBM PC standard.
As such, the development of the Android platform helped to accelerate the use of ARM processors throughout the industry. However, Intel processors still held a dominant market share in laptop devices due to the need for IBM PC compatible designs, and the mobile processors designed for these platforms were not suitable for mobile phones.
Fast forward to 2022, and it is clear that mobile computing is taking the lead. Laptops and desktop PCs are quickly being phased out in favour of devices that can fit in pockets, connect to external devices wirelessly, and cast to larger screens if needed. Not only are mobile devices becoming increasingly popular, large businesses (such as Apple) are now turning to custom processors that adopt either the ARM or RISC-V instruction set.
Not only is ARM starting to replace Intel in mainstream devices, it is even entering the server market with engineers preferring massive multicore RISC designs over CISC server cores that have more processing power per core, but far fewer cores. Furthermore, the rise of RISC-V is presenting engineers with new opportunities to create computing platforms that are open-source and independent from any licenses or up-marketed devices.
Overall, the CPU market is starting to fragment as new opportunities open themselves up to engineers, and Intel needs to consider changing tactic if it is to remain relevant going forward.
Recent reports from numerous sources suggest that Intel is looking into RISC-V as a new architecture to develop and deploy, and many believe that the adoption of RISC-V will be to provide customers with RISC-V cores. At the same time, development of RISC-V from Intel would also allow Intel to develop massive multicore designs with varying power capabilities (such as the Apple M1 which has both high-performance and high-efficiency cores).
However, RISC-V being a RISC type CPU provides Intel with a golden opportunity to enter the microcontroller market, something that Intel has virtually no presence in. While Intel has developed microcontrollers, the last microcontroller released was the Intel Quark Microcontroller back in 2018, and this was based around the x86 instruction set.
If Intel can leverage RISC-V into microcontrollers, it would be one of the few semiconductor manufacturers in the world who can design, develop, and manufacture its own parts (many other suppliers have to rely on third-party foundries). At the same time, Intel can take advantage of the years of experience that Intel has with designing high-performance microprocessors to potentially create some of the industries best RISC-V microcontrollers.
Considering that Intel already manufactures ASICs, FPGAs, chipsets, GPUs, and memory, it has the technology to create an entirely Intel solution that would be able to take full advantage of all hardware. Such SoCs could present engineers with a low cost, low power product able to be used in mobile platforms, support RISC-V software, and integrate numerous peripherals all designed to work together.
Despite the rise in ARM and RISC-V, it is highly unlikely that Intel would ever walk away from their long standing x86 ISA.
While ARM and RISC-V may be showing signs of use in servers, the Intel x86 ISA provides designer with a CISC architecture which provides more instructions capable of doing more things. Of course, RISC CPUs can simulate these functions by executing a series of simpler instructions, but this is not always desired. At the same time, the number of systems currently in use that depend on x86 practically guarantees that it will continued to be used for the next 50 years.
There are also plenty of legacy systems that cannot be easily replaced, and this will maintain the x86 market to some degree. Furthermore, improvements on these legacy systems will also create new demand for x86 devices as well as those who can write software targeted at x86 systems.
Overall, Intel has a solid market share on the worlds supply of CPUs, but the rising microcontroller and mobile processing markets present a potential threat to Intel if it doesn’t change its act. At the same time, Intel needs to recognise the technologies that is has and consider moving into new industries such as microcontrollers where it could dominate the industry.