On Thu, 16 Jan 2003, Quentin Mason wrote: > > Hi, > I was wondering whether it was common knowledge or whether anyone else has > noticed how much faster it is to compile things on MandrakePPC over an > Intel machine ? > For example: > 500 Mhz 1Mb L2 Cache, 384Mb RAM Wallstreet Powerlogix G3 upgrade, g++ 2.95 ~3mins > Dual 933 MHz Pentium 3, 1Gb RAM , g++ 3.2.1 2.25 >Hrs > > This is on a ~ 1Mb file that I was trying to compile for my research > without ANY optimisation options at all. The file was generated by some > other code and consists entirely of statements like > -1/3*sin(a+b+c)+2*sin(a-b+c). > > I discovered this by accident trying to do some work on my laptop at home > over Xmas; I thought it would be fun to try several of these huge object > files out on my laptop when they take so long on the machines at work. I > have some other files to compile that take >24 hrs w/ >900Mb RAM + >10Mb > stacksize in the office. I could imagine a factor of 2-3 difference, but > to have around 100 including the processor speeds was _very_ surprising. > > Does anyone have any experience to compare to ? >
My experience building the packages with the optimizations we use has been the opposite, but I don't really know what an equivalent x86 machine might be to make a fair comparison. Comparing my local builds with building the same package on x86 on the Paris cluster, it takes me as much as 10 times longer, in some cases. Stew Benedict -- MandrakeSoft PPC FAQ: http://www.linux-mandrake.com/en/ppcFAQ.php3 IRC: irc.openproject.net #cooker-ppc Archives: http://marc.theaimsgroup.com/?l=mandrake-cooker-ppc&r=1&w=2
