[normaljoe]

RC vs GC really depends on workload/memory and the volatility of the allocate to deallocate timeframe.

I think the better argument here is that iOS and MacOS use RC in the underlying objc libs. Having a CPU that works better around that makes sense to increase performance for those particular OSes.

[olliej]

Oh I didn’t mean to distract from the improvements to refcounting - it sounds very much like they’ve significantly improved the perf of uncontended atomic increment vs Intel which is obviously a win on iOS and OS X as objc and Swift both inline refcounting - I think on windows/COM it is through a virtual call? In which case improving the increment itself seems like it would not be a huge win.

As far as perf the general argument is that dropping the need for refcounting saves time, and that removal also helps caching due to reduced per-object size.

That said I’m not sure if those comparisons are comparing to generational or moving collectors (which are the low latency collectors) because those start needing write barriers.