A synthetic benchmarking tool to measure peak capabilities of a opencl device. It only measures the peak metrics that can be achieved using vector operations and does not represent a real-world use case
eg
Platform: AMD Accelerated Parallel Processing
Device: Tahiti
Driver version : 1445.5 (VM) (Linux x64)
Compute units : 32
Clock frequency : 1000 MHz
Global memory bandwidth (GBPS)
float : 211.27
float2 : 220.40
float4 : 217.36
float8 : 116.76
float16 : 59.32
Single-precision compute (GFLOPS)
float : 3700.45
float2 : 3548.87
float4 : 3531.60
float8 : 3528.15
float16 : 3518.72
Double-precision compute (GFLOPS)
double : 864.44
double2 : 860.07
double4 : 859.88
double8 : 850.53
double16 : 830.28
Integer compute (GIOPS)
int : 783.14
int2 : 783.54
int4 : 776.09
int8 : 753.42
int16 : 753.52
Transfer bandwidth (GBPS)
enqueueWriteBuffer : 8.86
enqueueReadBuffer : 5.29
enqueueMapBuffer(for read) : 22093.45
memcpy from mapped ptr : 3.99
enqueueUnmap(after write) : 43471.33
memcpy to mapped ptr : 3.92
Kernel launch latency : 38.14 us
Send in results of your device to [email protected]