3 ms·

Windows on ARM will never materialize because Qualcomm can't make compelling hardware for it to run on. Keep in mind that the M1 at launch was as fast as a lot

Windows on ARM will never materialize because Qualcomm can't make compelling hardware for it to run on. Keep in mind that the M1 at launch was as fast as a lot of Intel's lineup running emulated x86 apps. That's what Qualcomm needs to be able to pull off. If the performance is there then people will buy the laptops. That changes the developer story from "please recompile and retest your apps so they'll run on laptops nobody's buying" to "if you recompile your app it'll run 20% faster on this already screaming fast laptop chip".

jwells89 3y ago
Proficiency in emulating x86 is just one piece of the puzzle, though. It helped drive M1 adoption to be sure, but users also had a lot of confidence that developers would pull through and provide native binaries in a timely manner, catapulting the capabilities and battery life of their shiny new laptops even further. Apple and third party Mac devs had pulled it off twice already, so there was good reason to believe that they’d do it again. If the new Qualcomm chips have good performance, it’ll drive sales initially, but enthusiasm will fizzle if 2-3 years down the road developers aren’t making meaningful efforts to port their Windows versions to ARM, because a Windows on ARM device running mostly emulated processes is almost certainly going to fall behind M-series Macs running mostly native in performance and battery life. If Qualcomm can’t keep up substantial performance improvements on a yearly cadence, they also risk getting lapped by traditional x86 machines.
- jmholla 3y ago
  > but users also had a lot of confidence that developers would pull through and provide native binaries in a timely manner, catapulting the capabilities and battery life of their shiny new laptops even further. Yea. There was a real concerted effort by not Apple employees to get all these *modern* tools working on M1s. I remember at the company I was at, a small handful of engineers working diligently to get our internal dev software and processes working on M1s.

jwells89 3y ago

Proficiency in emulating x86 is just one piece of the puzzle, though. It helped drive M1 adoption to be sure, but users also had a lot of confidence that developers would pull through and provide native binaries in a timely manner, catapulting the capabilities and battery life of their shiny new laptops even further. Apple and third party Mac devs had pulled it off twice already, so there was good reason to believe that they’d do it again. If the new Qualcomm chips have good performance, it’ll drive sales initially, but enthusiasm will fizzle if 2-3 years down the road developers aren’t making meaningful efforts to port their Windows versions to ARM, because a Windows on ARM device running mostly emulated processes is almost certainly going to fall behind M-series Macs running mostly native in performance and battery life. If Qualcomm can’t keep up substantial performance improvements on a yearly cadence, they also risk getting lapped by traditional x86 machines.

jmholla 3y ago
> but users also had a lot of confidence that developers would pull through and provide native binaries in a timely manner, catapulting the capabilities and battery life of their shiny new laptops even further. Yea. There was a real concerted effort by not Apple employees to get all these *modern* tools working on M1s. I remember at the company I was at, a small handful of engineers working diligently to get our internal dev software and processes working on M1s.