PCTuning fórum

The first one is that there is a new stepping coming, SemiAccurate is hearing mid-to late Q3/2012 for the next rev. That revis said to bump performance, specifically integer performance, up by quite a bit, and possibly improve clocks too. Either way, it looks like that stepping is the one to keep an eye out for.It isn’t a Barcelona type fiasco, but it isn’t an HD4870 launch either.

flanker píše:krucinál, kde má tedy x87 ještě praktický smysl? Mě to jako overclockera jen mrzí z důvodu, že WR AMD asi v superpí ještě nepadne....

SSE instructions: The original AMD64 architecture adopted Intel's SSE and SSE2 as core instructions. SSE3 instructions were added in April 2005. SSE2 replaces the x87 instruction set's IEEE 80-bit precision with the choice of either IEEE 32-bit or 64-bit floating-point mathematics. This provides floating-point operations compatible with many other modern CPUs. The SSE and SSE2 instructions have also been extended to operate on the eight new XMM registers. SSE and SSE2 are available in 32-bit mode in modern x86 processors; however, if they're used in 32-bit programs, those programs will only work on systems with processors that have the feature. This is not an issue in 64-bit programs, as all AMD64 processors have SSE and SSE2, so using SSE and SSE2 instructions instead of x87 instructions does not reduce the set of machines on which x64 programs can be run. SSE and SSE2 are generally faster than, and duplicate most of the features of the traditional x87 instructions, MMX, and 3DNow!.

The default instruction set for code generation using the x86 VC++ 11.0 compiler is SSE2. This will be better documented closer to the release date. If you need to target older architectures you can use /arch:IA32.

DOC_ZENITH píše:Protože kdyby to bylo jinak, tak by se přeci každej soft jen prohnal x-compilery, existovalo by rázem x-verzí každá optimalizovaná pro svůj typ CPU a vše by všem chodilo krásně protože nám to compiler "sám na míru instrukčních sad jemu daných optimalizoval" ale ee, máme jednu verzi programu, v 90+% případů sedící na prvních verzích SSE nebo čistě x86+x87. Otázka zní teda proč?

nou píše:

DOC_ZENITH píše:Protože kdyby to bylo jinak, tak by se přeci každej soft jen prohnal x-compilery...

ciastocne to tak funguje. ale clovek tym dostane len skalarny SSE kod...

fact = 1;
for(int i=2;i<=10;i++)
  fact*=i;

v = (1,2);
fv = (1,2);
k = (2,2);
for(int i=1;i<=4;i++)
{
  v = v+k;
  fv = v.*fv;
}
fact = fv[0]*fv[1];

v = [1,2];
fv = [1,2];
k = 2;
for(int i=1;i<=4;i++)
{
  for(int j=0;j<2;j++)
     v[j] = v[j]+k;
  for(int j=0;j<2;j++)
     fv[j] = v[j]*fv[j];
}
fact = fv[0]*fv[1];