1 2Platform: NVIDIA CUDA 3 Device: Graphics Device 4 Driver version : 378.13 (Linux x64) 5 Compute units : 28 6 Clock frequency : 1683 MHz 7 8 Global memory bandwidth (GBPS) 9 float : 389.99 10 float2 : 394.86 11 float4 : 410.15 12 float8 : 388.05 13 float16 : 263.58 14 15 Single-precision compute (GFLOPS) 16 float : 11675.87 17 float2 : 13240.07 18 float4 : 13317.21 19 float8 : 13151.05 20 float16 : 12939.08 21 22 Double-precision compute (GFLOPS) 23 double : 425.21 24 double2 : 432.63 25 double4 : 425.45 26 double8 : 420.62 27 double16 : 409.39 28 29 Integer compute (GIOPS) 30 int : 3507.68 31 int2 : 3801.87 32 int4 : 3772.84 33 int8 : 3774.45 34 int16 : 3748.59 35 36 Transfer bandwidth (GBPS) 37 enqueueWriteBuffer : 9.96 38 enqueueReadBuffer : 8.95 39 enqueueMapBuffer(for read) : 11.11 40 memcpy from mapped ptr : 12.16 41 enqueueUnmap(after write) : 12.40 42 memcpy to mapped ptr : 12.48 43 44 Kernel launch latency : 4.22 us 45 46