RE: Benchmarks
Oops. I took a look at your page of results, and noticed that the OMAP2420 didn't really perform all that well at floating point tests, despite its having a hardware VFP (vector floating point) unit. It then dawned on me that although I'd compiled nbench with hardware fp, it had been linked against a soft-float libm. If nothing else, comparing the differences between these new results and the ones I previously posted shows how much of a difference the floating point libm makes. So here are the new results with everything hardware floating point (I've also flashed an updated image since I posted the other results, but that should make little difference): Nokia-N800-26:~# ./nbench BYTEmark* Native Mode Benchmark ver. 2 (10/95) Index-split by Andrew D. Balsa (11/97) Linux/Unix* port by Uwe F. Mayer (12/96,11/97) TEST: Iterations/sec. : Old Index : New Index : : Pentium 90* : AMD K6/233* :--:-: NUMERIC SORT: 116.56 : 2.99 : 0.98 STRING SORT : 6.5519 : 2.93 : 0.45 BITFIELD: 3.2551e+07 : 5.58 : 1.17 FP EMULATION: 13.204 : 6.34 : 1.46 FOURIER : 996.52 : 1.13 : 0.64 ASSIGNMENT : 1.0108 : 3.85 : 1.00 IDEA: 284.63 : 4.35 : 1.29 HUFFMAN : 119.99 : 3.33 : 1.06 NEURAL NET : 1.1723 : 1.88 : 0.79 LU DECOMPOSITION: 39.681 : 2.06 : 1.48 ==ORIGINAL BYTEMARK RESULTS== INTEGER INDEX : 4.029 FLOATING-POINT INDEX: 1.637 Baseline (MSDOS*) : Pentium* 90, 256 KB L2-cache, Watcom* compiler 10.0 ==LINUX DATA BELOW=== CPU : L2 Cache: OS : Linux 2.6.18-omap1 C compiler : gcc version 3.4.4 (release) (CodeSourcery ARM 2005q3-2) libc: ld-2.3.6.so MEMORY INDEX: 0.808 INTEGER INDEX : 1.185 FLOATING-POINT INDEX: 0.908 Baseline (LINUX): AMD K6/233*, 512 KB L2-cache, gcc 2.7.2.3, libc-5.4.38 * Trademarks are property of their respective holder. Cheers, Simon ___ OpenMoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Re: Benchmarks
I don't own a Neo yet, but I have here a Samsung S3C2440A (arm920 based), and here are the results : BYTEmark* Native Mode Benchmark ver. 2 (10/95) Index-split by Andrew D. Balsa (11/97) Linux/Unix* port by Uwe F. Mayer (12/96,11/97) TEST: Iterations/sec. : Old Index : New Index : : Pentium 90* : AMD K6/233* :--:-: NUMERIC SORT: 91.523 : 2.35 : 0.77 STRING SORT : 8.307 : 3.71 : 0.57 BITFIELD: 3.0876e+07 : 5.30 : 1.11 FP EMULATION: 7.9239 : 3.80 : 0.88 FOURIER : 11.727 : 0.01 : 0.01 ASSIGNMENT : 0.48694 : 1.85 : 0.48 IDEA: 340.46 : 5.21 : 1.55 HUFFMAN : 40.497 : 1.12 : 0.36 NEURAL NET :0.014012 : 0.02 : 0.01 LU DECOMPOSITION: 0.48278 : 0.03 : 0.02 ==ORIGINAL BYTEMARK RESULTS== INTEGER INDEX : 2.941 FLOATING-POINT INDEX: 0.020 Baseline (MSDOS*) : Pentium* 90, 256 KB L2-cache, Watcom* compiler 10.0 ==LINUX DATA BELOW=== CPU : L2 Cache: OS : Linux 2.6.13-rc7 C compiler : gcc version 3.4.6 libc: libc-2.3.6.so MEMORY INDEX: 0.673 INTEGER INDEX : 0.783 FLOATING-POINT INDEX: 0.011 Baseline (LINUX): AMD K6/233*, 512 KB L2-cache, gcc 2.7.2.3, libc-5.4.38 * Trademarks are property of their respective holder. ___ OpenMoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Benchmarks
I presume the OMAP processors are comparable and of interest. Here are the results for the OMAP1710 (Nokia 770) and OMAP2420 (Nokia N800): Nokia 770 OMAP1710 Results == CFLAGS = -s -static -O3 -fomit-frame-pointer -Wall \ -fforce-addr -fforce-mem -falign-loops=2 -falign-functions=2 \ -falign-jumps=2 -funroll-loops BYTEmark* Native Mode Benchmark ver. 2 (10/95) Index-split by Andrew D. Balsa (11/97) Linux/Unix* port by Uwe F. Mayer (12/96,11/97) TEST: Iterations/sec. : Old Index : New Index : : Pentium 90* : AMD K6/233* :--:-: NUMERIC SORT: 75.819 : 1.94 : 0.64 STRING SORT : 5.1752 : 2.31 : 0.36 BITFIELD: 2.4423e+07 : 4.19 : 0.88 FP EMULATION: 9.603 : 4.61 : 1.06 FOURIER : 68.432 : 0.08 : 0.04 ASSIGNMENT : 0.7025 : 2.67 : 0.69 IDEA: 224.89 : 3.44 : 1.02 HUFFMAN : 89.866 : 2.49 : 0.80 NEURAL NET :0.092834 : 0.15 : 0.06 LU DECOMPOSITION: 3.0535 : 0.16 : 0.11 ==ORIGINAL BYTEMARK RESULTS== INTEGER INDEX : 2.960 FLOATING-POINT INDEX: 0.122 Baseline (MSDOS*) : Pentium* 90, 256 KB L2-cache, Watcom* compiler 10.0 ==LINUX DATA BELOW=== CPU : L2 Cache: OS : Linux 2.6.16.27-omap1 C compiler : gcc version 3.4.4 (release) (CodeSourcery ARM 2005q3-2) libc: ld-2.3.6.so MEMORY INDEX: 0.601 INTEGER INDEX : 0.862 FLOATING-POINT INDEX: 0.068 Baseline (LINUX): AMD K6/233*, 512 KB L2-cache, gcc 2.7.2.3, libc-5.4.38 * Trademarks are property of their respective holder. Nokia N800 OMAP2420 Results === CFLAGS = -s -static -O3 -fomit-frame-pointer -Wall \ -fforce-addr -fforce-mem -falign-loops=2 -falign-functions=2 \ -falign-jumps=2 -funroll-loops -mfpu=vfp -mfloat-abi=softfp BYTEmark* Native Mode Benchmark ver. 2 (10/95) Index-split by Andrew D. Balsa (11/97) Linux/Unix* port by Uwe F. Mayer (12/96,11/97) TEST: Iterations/sec. : Old Index : New Index : : Pentium 90* : AMD K6/233* :--:-: NUMERIC SORT: 113.55 : 2.91 : 0.96 STRING SORT : 6.6113 : 2.95 : 0.46 BITFIELD: 3.2539e+07 : 5.58 : 1.17 FP EMULATION: 13.183 : 6.33 : 1.46 FOURIER : 103.04 : 0.12 : 0.07 ASSIGNMENT : 1.0084 : 3.84 : 1.00 IDEA: 283.98 : 4.34 : 1.29 HUFFMAN : 120.01 : 3.33 : 1.06 NEURAL NET : 0.50151 : 0.81 : 0.34 LU DECOMPOSITION: 39.306 : 2.04 : 1.47 ==ORIGINAL BYTEMARK RESULTS== INTEGER INDEX : 4.016 FLOATING-POINT INDEX: 0.577 Baseline (MSDOS*) : Pentium* 90, 256 KB L2-cache, Watcom* compiler 10.0 ==LINUX DATA BELOW=== CPU : L2 Cache: OS : Linux 2.6.18-omap1 C compiler : gcc version 3.4.4 (release) (CodeSourcery ARM 2005q3-2) libc: ld-2.3.6.so MEMORY INDEX: 0.810 INTEGER INDEX : 1.176 FLOATING-POINT INDEX: 0.320 Baseline (LINUX): AMD K6/233*, 512 KB L2-cache, gcc 2.7.2.3, libc-5.4.38 * Trademarks are property of their respective holder. Cheers, Simon ___ OpenMoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community
Benchmarks
Hello all, Over at the SVHMPC, we have been benchmarking application processors. The main drive for this is that I have been working on a Freescale i.MX31 processor (based on an ARM11 core) and wanted to see how it performs compared to the competition. So far we have benchmarks for a 400Mhz PXA255, 600Mhz PXA270 and 530Mhz i.MX31. I would very much like to see how the S3C2410 and SC32442 used in the GTA01 and GTA02 stack up against the others. We may also be getting results from someone with a PXA320. If you're interested, could someone with a GTA01 (and GTA02 if possible) please compile run the benchmark? We're using the nbench software as it's easy to cross-compile. nbench is available at http://www.tux.org/~mayer/linux/bmark.html. I've posted the results we have so far on our wiki, http://hbmobile.org/wiki/index.php?title=Application_Processor_Benchmark s These are preliminary results and currently do not match what others have found, so some tweaking might be needed (especially on the i.MX31!!!). I look forward to adding more results as they come in. Cheers, Tom ___ OpenMoko community mailing list community@lists.openmoko.org http://lists.openmoko.org/mailman/listinfo/community