
Starting from zero
I'd suggest Qualcomm trying to get into the DC server market at this point in time is perhaps rather late in the hype cycle?
How long will it take to create a profitable business of credible scale from scratch? The progress in PC CPU is a fair base to work from, but even so I'm guessing six years to get reasonable scale and a credible ROI. That's a long time when everybody is already building out DCs in huge volume, and the real money is in GPUs. If Qualcomm can really add something new like a step change in energy use per instruction, then perhaps, but with everybody else already in this market and chasing that goal, who'd bet on Qualcomm?