3 ms·
They actually did this in the M1 for retain/release for Objective-C and Swift objects. It's an extremely common operation on the platform and is 4x faster on M1
by simonh 6y ago
They actually did this in the M1 for retain/release for Objective-C and Swift objects. It's an extremely common operation on the platform and is 4x faster on M1 natively, and still 2x faster than native Intel under Rosetta.