On Saturday 03 February 2007 01:02:20 Jason Chu wrote: > On Fri, 2 Feb 2007 16:21:08 -0800 (PST) > > [EMAIL PROTECTED] wrote: > > > Most benchmarks I have seen show no real performance gain with > > > x64. If the > > > piece of software in question is built and optimized for x64, then > > > we'll see > > > gains, but not much software is in that boat right now. > > > > The biggest gains come with number crunching. I'm working with > > cryptographics hash functions where I need to multiply large numbers > > together. A 64-bit multiply is able to do 4 times the work of a > > 32-bit multiply, not to mention the overhead involved with assembling > > the 32-bit pieces together. This can lead to huge performance gains, > > but only because it is tailored to such an improvement. Most GUI > > programs are IO bound and can't really take advantage of 64-bits. > > > > -Joe > > That and accessing more than 4GB of ram. The address space of a 32-bit > processor is limited by the size of a 32-bit register. With 64-bit > registers, your address space is that much larger. > > Jason
That's the biggest gain of them all I must say, seemless large arse pointers -- Guillermo A. Amaral, BCSE @ site: http://www.guillermoamaral.com/ @ blog: http://blog.guillermoamaral.com/ # nick: thewonka / thewonka81
pgprmQgvOvPBv.pgp
Description: PGP signature
_______________________________________________ arch mailing list arch@archlinux.org http://www.archlinux.org/mailman/listinfo/arch
