All, Java, C#, & C, tests are on Windows XP running under Parallels on a Mac Book Pro., 2.33 GHz Intel Core 2 Duo, 2 GB 667 MHz DDR2 SDRAM.
I also ran the large set using gcc compiled with -O3 on g++ (GCC) 3.4.5 (mingw special), the results were: SciGMark 1.0 - C++ - specialized Using 2.00 seconds min time per kenel. Composite Score: 385.09 FFT Mflops: 71.93 (N=1048576) SOR Mflops: 662.46 (1000 x 1000) MonteCarlo: Mflops: 90.57 Sparse matmult Mflops: 452.50 (N=100000, nz=1000000) LU Mflops: 564.25 (M=1000, N=1000) MultPoly Mflops: 71.93 (N=40) Which makes Java slightly faster than GCC. So you have to give kudos to John Rose and Co. at Sun, they have produced a fast VM. Were is the mistake in the C code, I couldn't find it? On Apr 25, 11:05 am, hlovatt <[EMAIL PROTECTED]> wrote: > Your results seem to be the outliers - are you certain you are using > the Scimarks referenced in the paper? > > When I compile the C# benchmarks with: > > c:\WINDOWS\Microsoft.NET\Framework\v3.5\csc /optimize+ /debug- / > checked- *.cs > > And run the large set with: > > commandline -large > > I get: > > SciGMark 1.0 - C# - specialized > > FFT (1048576): 68.24358711148 > SOR (1000x1000): 336.610476965462 > Monte Carlo : 103.895055082435 > Sparse matmult (N=100000, nz=1000000): 363.889491977595 > LU (1000x1000): 472.132470273794 > PolyMult (N=40): 282.144700451176 > Composite Score: 271.152630310324 > > Platform Information > CLR Version: 2.0.50727.1433 > Working Set: 21704704 > > When I compile the Java code with no compiler options and use the > server VM I get: > > SciGMark 1.0 - Java - specialized > FFT (1048576): 66.14391779366471 > SOR (1000x1000): 862.4075989952408 > Monte Carlo : 210.20788286956355 > Sparse matmult (N=100000, nz=1000000): 308.99216463285444 > LU (1000x1000): 512.0327520949826 > PolyMult (N=100): 432.48570135288116 > > Composite Score: 398.7116696231979 > > java.vendor: Sun Microsystems Inc. > java.version: 1.6.0_06 > os.arch: x86 > os.name: Windows XP > os.version: 5.1 > > Which makes Java about 1.6 times quicker (essentially the same as the > small set results). > > Your results don't seem to tie up with other peoples - are you certain > you are running the right benchmarks? The claim that Java is faster > than C# seems quite reasonable given this set of results - why are > yours different than other peoples? > > On Apr 24, 4:34 am, Jon Harrop <[EMAIL PROTECTED]> wrote: > > > On Wednesday 23 April 2008 19:11:08 Thomas E Enebo wrote: > > > > Jon, Can you specify how you are running these benchmarks? > > > For C++ and Java, I downloaded the source, compiled it with "make" and ran > > it. > > For C#, I copied the source into a VS project, set it to "Release" mode, > > built and ran it. > > > > I did not see version of arguments to JVM (or .Net runtime). > > > Neither did I. :-) > > > > You are using -server I assume... > > > Yes. That is the default here. > > > -- > > Dr Jon D Harrop, Flying Frog Consultancy > > Ltd.http://www.ffconsultancy.com/products/?e --~--~---------~--~----~------------~-------~--~----~ You received this message because you are subscribed to the Google Groups "JVM Languages" group. To post to this group, send email to [email protected] To unsubscribe from this group, send email to [EMAIL PROTECTED] For more options, visit this group at http://groups.google.com/group/jvm-languages?hl=en -~----------~----~----~----~------~----~------~--~---
