wgblas
    Preparing search index...

    Module benchmarks/nvidia-geforce-gtx-1650/idamax

    Benchmark results for idamax on Nvidia Geforce Gtx 1650.

    n wgblas ms wgblas GB/s cuBLAS ms cuBLAS GB/s efficiency
    32 0.0212 0.0120 0.0168 0.0152 79.3%
    64 0.0214 0.0239 0.0176 0.0291 82.1%
    128 0.0208 0.0493 0.0197 0.0519 95.0%
    512 0.0211 0.1944 0.0170 0.2408 80.7%
    1024 0.0214 0.3832 0.0169 0.4848 79.0%
    4096 0.0216 1.5204 0.0181 1.8060 84.2%
    16384 0.0219 5.9796 0.0179 7.3143 81.8%
    65536 0.0235 22.3520 0.0263 19.9562 112.0%
    262144 0.0363 57.7410 0.0386 54.2741 106.4%
    1048576 0.0700 119.8099 0.0839 99.9977 119.8%
    4194304 0.2078 161.4436 0.2666 125.8568 128.3%
    16777216 0.7641 175.6482 1.0618 126.4068 139.0%

    Efficiency = wgblas GB/s ÷ cuBLAS GB/s × 100. 100% means parity with cuBLAS; values above 100% mean wgblas achieved greater throughput.

    idamax-default GB/s chart

    idamax-default ms chart

    Unless noted otherwise, every result above uses unit stride (incx = incy = 1) — the normal case, and the coalesced, best-case GPU access pattern. Real usage sometimes passes a non-unit stride (e.g. operating on a row or column of a larger matrix, where incx = lda), which breaks memory coalescing and costs measurably more. This section sweeps a few representative strides to characterize that cost separately, collapsed below by default — expand a stride to see its table and chart.

    Nvidia Geforce Gtx 1650 — stride = 4
    n wgblas ms wgblas GB/s cuBLAS ms cuBLAS GB/s efficiency
    32 0.0212 0.0121 0.0163 0.0157 76.9%
    64 0.0208 0.0246 0.0169 0.0302 81.4%
    128 0.0206 0.0497 0.0171 0.0599 83.0%
    512 0.0208 0.1968 0.0167 0.2457 80.1%
    1024 0.0210 0.3908 0.0168 0.4876 80.2%
    4096 0.0216 1.5170 0.0181 1.8124 83.7%
    16384 0.0225 5.8182 0.0180 7.2753 80.0%
    65536 0.0349 15.0105 0.0306 17.1381 87.6%
    262144 0.0696 30.1176 0.0640 32.7762 91.9%
    1048576 0.2096 40.0281 0.1999 41.9699 95.4%
    4194304 0.7802 43.0053 0.7362 45.5764 94.4%

    idamax-stride4 GB/s chart

    idamax-stride4 ms chart

    Nvidia Geforce Gtx 1650 — stride = 32
    n wgblas ms wgblas GB/s cuBLAS ms cuBLAS GB/s efficiency
    32 0.0192 0.0134 0.0167 0.0153 87.3%
    64 0.0191 0.0268 0.0139 0.0368 72.9%
    128 0.0188 0.0546 0.0172 0.0595 91.7%
    512 0.0195 0.2104 0.0170 0.2411 87.2%
    1024 0.0197 0.4156 0.0164 0.4995 83.2%
    4096 0.0228 1.4402 0.0174 1.8806 76.6%
    16384 0.0331 3.9632 0.0216 6.0592 65.4%
    65536 0.0661 7.9341 0.0395 13.2664 59.8%
    262144 0.2023 10.3672 0.1078 19.4584 53.3%

    idamax-stride32 GB/s chart

    idamax-stride32 ms chart

    Nvidia Geforce Gtx 1650 — stride = 256
    n wgblas ms wgblas GB/s cuBLAS ms cuBLAS GB/s efficiency
    512 0.0196 0.2088 0.0162 0.2522 82.8%
    1024 0.0197 0.4166 0.0164 0.5010 83.2%
    4096 0.0290 1.1284 0.0180 1.8204 62.0%
    16384 0.0449 2.9205 0.0294 4.4619 65.5%
    65536 0.1021 5.1328 0.0501 10.4556 49.1%

    idamax-stride256 GB/s chart

    idamax-stride256 ms chart

    See also:

    Settings
    Member Visibility

    On This Page
    Nvidia Geforce Gtx 1650 — wgblas vs cuBLASSee alsoStride sweep