From the original SynthMaker forum (2004–2014), with its members’ names and dates.
infuzion ·
I suggested before, but even now I can see I might want to optimize my ASM using 2 different methods for AMDx64 & Core each. Not sure the best way; maybe an Int output for the CPU, or a float (AMD x64 = 1.0, Optron=1.1 P3=2.0, P4=3.0, P4E=3.1...), a text output... I vote for Int out to be used easier with a Select.
TIA
infuzion ·
I expanded the request to cover the # cores to cover Signal Analyser bug. cheers
infuzion ·
Here's some ASM code I found around to help out, prob can be reduced by dropping <SSE-CPU support: http://www.coli.uni-saarland.de/~eric/s ... vel2007.zi p MyCo's Advanced Reverb could use such a primitive.
TomC ·
infuzion wrote: I vote for Int out to be used easier with a Select.
TIA
+1
One change: please not one single primitve (will become messy rather soon).
1. general CPUID 2. SSE level in increasing order (0=no SSE). Increasing because I'd like to have my old coding still work with newer CPUs. 3. Number of cores (CPU independent). If number of CPUs is possible it would nice to have, too.
Tom
infuzion ·
TomC wrote: One change: please not one single primitive (will become messy rather soon). 1. general CPUID 2. SSE level in increasing order (0=no SSE). Increasing because I'd like to have my old coding still work with newer CPUs. 3. Number of cores (CPU independent). If number of CPUs is possible it would nice to have, too.
Well, with just a CPUID, the SSE level should be able to be found via a lookup table in a module, likely as the # of cores (unless someone disabled the extra cores in BIOS).
Does anyone know if different CPUs share the same CPUID, but have different # of cores/SSE support?
Wrote something here? To have a post removed or your name hidden, write to [email protected].
SynthMaker