Hacker News new | ask | show | jobs
by jamie_ca 621 days ago
I was comparing notes with a new coworker the other week - I've got an M1 pro, he's got an M3. Running an individual test for a feature we were working on (in Rails) lines up with the 25-30% numbers quoted up thread a bit, but that's the difference between my machine taking 40sec or so vs his taking 30 or less.

It's noticeable when you've got it to compare against, and I'm looking forward to work's hardware refresh cycle bumping me to M4 next year, but I still don't think I'd upgrade if it were a personal machine (and if I can get a decent discount to buy out the depreciated M1 for personal use I'm planning on that rather than looking at anything new).

1 comments

That's my take as well. I used an M1 and an M3 for the same work, the M3 was definitely faster but the M1 wasn't bad. Both were substantially better than the Intel Macs I was using before (for the same tasks as well). Times cut from 5-10 minutes down to 1-2 minutes on the M1, and another 10-30 seconds shaved off with the M3. So faster, but not as game changing as the M1 itself.

The one thing I do like with the M3 vs M1 is when I had to, for reasons, run an x64 VM. I felt it was barely usable on the M1, but it was tolerable on the M3. The performance on the M3 of an x64 VM was close to the old Intel Macs I'd ditched, which were acceptable but hardly great. The M1 running a VM felt like a time warp back to my college days in the early '00s.