RISC-V: Did Apple Make the Wrong Choice?
Apple’s M1 chip is based on ARM. But what if the open source RISC-V instruction-set proves superior to ARM?
RISC-V is a competing microprocessor design. Or to be more precise it is a competing instruction set architecture (ISA). Some speculate that Apple may have made the wrong choice and that RISC-V is the future. Again to quote my sparring partner Sergiy Yevtushenko:
Worse (for Apple) is that wider acceptance of ARM for desktop opens the door to other non-x86 architectures as well. For example RISC-V recently started gaining attention. Unlike ARM it’s completely open source architecture which doesn’t need to be even licensed. And it has even lower power consumption per MHz than ARM, so high performance chips are much easier to design and use.
In fact this would be quite exciting in my humble opinion. RISC-V is a very interesting design which I would love to see more success from. However I will question the assumption that RISC-V somehow spells doom for Apple. Quite the contrary. Let me elaborate:
Apple has made more major hardware and operating system transitions than any other company in the computer industry:
- From Apple II to the first Mac in 1984. They successfully changed to an entirely different operating system and hardware architecture. They went from Apple DOS to MacOS. They went from 6502 CPUs to Motorola 68000 CPUs.
- In 1994 Apple transitioned again, this time from Motorola 6800x0 architecture to the PowerPC RISC processor made by IBM and Motorola.
- MacOS 9 to MacOS X. In 2001 they made this transition, suprisingly smoothly. The Microsoft transition from Windows 3.11 to Windows 95 was a far more bumpy ride. The 16-bit Windows 3.11 applications did not run in an isolated environment allowing them e.g. to kill the whole operating system if their crashed.
- Few years later in 2006 we got the famous PowerPC to Intel transition.
- Now they are doing an Intel to ARM transition which thus far looks like it will go very well.
Making transitions is basically an Apple speciality. It is what they do. It is what they know how to do better than anybody else. And that ability stems in large part from their complete vertical integration. Full control over hardware and software, means that if Apple changes hardware they can also make sure their software is modified to fit this change.
That means if RISC-V becomes the best ISA to build a computer around, who is better equipped to transition to this new architecture? Dell and HP? Of course not. Apple will have an advantage over the whole rest of the PC industry in that they could make such a transition happen more quickly than anybody.
The rest of the industry benefit more from a stable eco-system. If the hardware and software stack is stable, it is much easier for the PC industry to compete with Apple, as they can all keep optimizing their speciality to the max, whether that is making chips, graphics cards, or software.
A world with frequent paradigm shifts is different. In such a world, changes have to be perfectly coordinated. Whomever enjoys full vertical integration will have the upper hand.