On Jul 2, 2012, at 4:34 PM, Kyle Sluder wrote: > On Jul 2, 2012, at 4:32 PM, Jens Alfke wrote: > >> >> On Jul 2, 2012, at 4:17 PM, Kyle Sluder wrote: >> >>>> It depends. 64-bit values are twice as big as 32-bit ones, so they use up >>>> twice as much L2 cache and RAM. >>> >>> I would be surprised if cache is managed at anything other than multiples >>> of register width (64 bits). >> >> That's not the point. Data containing 64-bit values (in objects, structs, >> stack frames…) is obviously bigger than data containing smaller values. > > Not necessarily. The CPU could just mask off the top 32 bits when executing > 32-bit opcodes on in-cache data. Each 32-bit value is still taking up 64 bits > of cache space. That's probably a lot easier and more efficient than making > it possible to address arbitrary 32-bit slices of cache. > > Note that I'm only referring to cache here, not RAM.
I suspect you are both talking past each other. Jens' assertion is that if you had a 128 byte cache, you could store either 8 64-bit integers or 16 32-bit integers in it. Whereas Kyle is asserting that the CPU need only read 32-bits at a time (or less) from the cache for opcodes that deal with 32-bits (or less) of data at a time. Your both correct, but your looking at different parts of the same problem. -- David Duncan _______________________________________________ Cocoa-dev mailing list (Cocoa-dev@lists.apple.com) Please do not post admin requests or moderator comments to the list. Contact the moderators at cocoa-dev-admins(at)lists.apple.com Help/Unsubscribe/Update your Subscription: https://lists.apple.com/mailman/options/cocoa-dev/archive%40mail-archive.com This email sent to arch...@mail-archive.com
