So is your theory that everyone who doesn’t get to buy Arm (the company) will switch to RISC-V, or is nVidia special?
Not really, I oversimplified for brevity. Let's take a few scenarios.
ARM stays independent:
Slowly RISC-V vendors and in-house implementations rise up to compete against ARM. I know of multiple high volume users of M and R range CPUs working towards replacing those with RISC-V. I've read about other efforts to go after the A market. I think it's just a matter of time before this erodes ARMs value proposition and creates downward pricing pressure. It will be slow, as RISC-V is still missing pieces and inertia has it's value. But a stagnating market isn't encouraging. The longer ARM is independent, the more likely some RISC-V vendor comes along to compete against the A series. Eventually ARM will be smaller and less profitable, to survive they will either find news ways to add additional value, or they will be sold.
Nvidia has to decide if they will stay an ARM customer or grow RISC-V. That probably depends on Microsoft. Tough call.
Which brings us to ARM is acquired by NVidia:
I feel Nvidia's angle is to try and crack the market for PCs plus consoles, by building an M3 Pro Ultra competitor. They will invest heavily, trying displace Intel in PCs, and AMD in consoles. On the M and R licensing business side, they will milk it while they can, but they won't invest. They will try bundling their GPUs with A series contracts, possibly no longer investing in versions that don't include a GPU.
This won't be received well by Qualcomm on the A series side, or anyone on the M and R series side, and will accelerate industry efforts to move to RISC-V. Further confirming Nvidia's decision not to invest in those markets. The M and R cpus will eventually be open sourced in some way, and Nvidia will exit those markets completely. The A+GPU business will live on, but Qualcomm will attack it with a RISC-V or some other solution. Nvidia will make some money in the portable market with A+GPU, but the real money will be in Datacenter, plus some gravy from console, and PC. The portable market will always be at risk of being cut.
How about Qualcomm buys ARM?
This is more of a mixed bag. Qualcomm would probably treat the embedded market better, but they would apply the bundling tactics with the CPU and their cellular solution, which will piss-off the A series customers. Nvidia embraces RISC-V out of spite, possibly accelerating ARMs decline.
Summary:
In the first scenario, ARM slowly becomes a small player. In the second scenario, ARM displaces a large portion of Intel's business, but shrinks in mobile. In the third, it's tougher to predict, but ARM lives on in all 3 scenarios, even though it's role changes a lot.