1Platform: Portable Computing Language 2 Device: pthread-AMD EPYC 7A53 64-Core Processor 3 Driver version : 1.8 (Linux x64) 4 Compute units : 128 5 Clock frequency : 2000 MHz 664 warnings generated. 7 8 Global memory bandwidth (GBPS) 9 float : 30.74 10 float2 : 31.35 11 float4 : 31.11 12 float8 : 33.88 13 float16 : 27.41 14 15 Single-precision compute (GFLOPS) 16 float : 81.79 17 float2 : 160.46 18 float4 : 308.39 19 float8 : 589.43 20 float16 : 107.23 21 22 No half precision support! Skipped 23 24 Double-precision compute (GFLOPS) 25 double : 81.46 26 double2 : 159.08 27 double4 : 293.62 28 double8 : 55.83 29 double16 : 78.10 30 31 Integer compute (GIOPS) 32 int : 198.51 33 int2 : 378.71 34 int4 : 750.36 35 int8 : 1407.42 36 int16 : 2162.47 37 38 Integer compute Fast 24bit (GIOPS) 39 int : 122.67 40 int2 : 176.17 41 int4 : 344.37 42 int8 : 617.54 43 int16 : 113.10 44 45 Transfer bandwidth (GBPS) 46 enqueueWriteBuffer : 16.47 47 enqueueReadBuffer : 16.79 48 enqueueWriteBuffer non-blocking : 18.28 49 enqueueReadBuffer non-blocking : 18.62 50 enqueueMapBuffer(for read) : 5142.44 51 memcpy from mapped ptr : 15.46 52 enqueueUnmap(after write) : 6599.52 53 memcpy to mapped ptr : 20.98 54 55 Kernel launch latency : 107.54 us 56 57