2 ms·
If you check ARC code in a profiler, there's a shocking amount of time spent in the retain/release calls: https://floooh.github.io/2016/01/14/metal-arc.html ht
by flohofwoe 5y ago
If you check ARC code in a profiler, there's a shocking amount of time spent in the retain/release calls:
https://floooh.github.io/2016/01/14/metal-arc.html https://floooh.github.io/2016/01/14/metal-arc.html
There are ways around this overhead at least in Metal, but this requires at least as much effort as not using ARC to begin with.
- Twisol 5y agoI think the M1 CPU lowers some of the ARC retain/release to silicon -- it doesn't remove the problem, but it does seem to reduce the relative overhead significantly. https://news.ycombinator.com/item?id=25203924 https://news.ycombinator.com/item?id=25203924 > this requires at least as much effort as not using ARC to begin with. Designing a brand new CPU architecture certainly counts as "at least as much effort", yes. ^_^ P.S. While I'm here, your "handles are the better pointers" blog post is one of my all-time favorites. I appreciate you sharing your experiences!