Ask HN: Apple Silicon (M1 Pro) vs. Ryzen 9 7940HS (Gaming Laptop)

3 points by apfsx 2y ago ↗ HN

Has anyone transitioned from a Linux/Windows laptop to Apple Silicon as their main machine and could share their experiences? I’d really be interested in hearing any compatibility issues, troubles, or annoyances you’ve had. Particularly with your dev environment, Docker containers (x86), or building things that need to be deployed on x86 servers.

I’ve been using a 2020 ASUS G14 (Ryzen 9 4900HS / RTX 2060) as my main machine for the past 3 years. WSL2, Docker, the catalogue of Windows apps you can run from years ago has been pretty good to me but not without the occasional blue screens, loud fans, temperatures hitting 90C, etc. Considering a 16 inch MBP.

3 comments

[ 4.9 ms ] story [ 12.6 ms ] thread

smoldesu 2y ago ↗

It's not going to be a straightforward choice, at least in my experience. Try both, see which works better.

However, I will say that you're not going to need a gaming laptop to get a good experience with Docker. Cheapo dual-core laptops from a decade ago are good enough for sub-30c idle temps; you might be able to fix your current situation by disabling your dGPU and trying Linux/Docker natively.

apfsx 2y ago ↗

I can definitely get 30C idle temps if I turn off the dGPU but I use it most of the time docked up to external monitors which forces the dGPU to run. At that point with my usual programs open I'm at 65-90C depending on what I'm doing within those programs. I can get this down to 50-55C stable if I disable boost on the CPU which keeps it at 3GHz no turbo but there is a noticeably performance hit.

IronWolve 2y ago ↗

Doable, but performance issues. If your going to run x86 containers will run slower and with more issues. And macs also crash when running lots of vm's like windows.

I'd say if your doing x86 dev, stay on amd. And then you get the bonus of the nvidia card for cuda for faster AI.

Or just use the mac as a front end dev, and put your containers in the cloud, and use the cloud for AI. More expensive, but get work to pay for it, or expense it.