Difference between revisions of "Rockpi4/dev/install-opencl"
(→Install libmali package) |
RadxaYuntian (Talk | contribs) (→Sanity Checkout with clpeak) |
||
| Line 16: | Line 16: | ||
To see if the opencl driver works properly on ROCK Pi 4, use the benchmark tool '''clpeak''' | To see if the opencl driver works properly on ROCK Pi 4, use the benchmark tool '''clpeak''' | ||
| − | $ sudo apt-get update && sudo apt-get install cmake git | + | $ sudo apt-get update && sudo apt-get install cmake git g++ |
$ git clone https://github.com/krrishnarraj/clpeak | $ git clone https://github.com/krrishnarraj/clpeak | ||
| − | $ cd clpeak | + | $ mkdir clpeak/build |
| − | $ cmake . | + | $ cd clpeak/build |
| − | $ make | + | $ cmake .. |
| + | $ make -j$(nproc) | ||
$ ./clpeak | $ ./clpeak | ||
Latest revision as of 04:20, 16 March 2023
ROCK Pi 4 > Development > Install OpenCL
Contents
OpenCL driver works on ROCK Pi 4
Install libmali package
Install libmali package from Radxa APT's testing repo.
$ sudo apt-get install rockchip-mali-midgard14 $ sudo apt-get install rockchip-mali-midgard-dev
Sanity Checkout with clpeak
To see if the opencl driver works properly on ROCK Pi 4, use the benchmark tool clpeak
$ sudo apt-get update && sudo apt-get install cmake git g++ $ git clone https://github.com/krrishnarraj/clpeak $ mkdir clpeak/build $ cd clpeak/build $ cmake .. $ make -j$(nproc) $ ./clpeak
After executing command clpeak, there shall be the following outputs:
Platform: ARM Platform
Device: Mali-T860
Driver version : 1.2 (Linux ARM64)
Compute units : 4
Clock frequency : 200 MHz
Global memory bandwidth (GBPS)
float : 3.68
float2 : 5.95
float4 : 6.70
float8 : 5.59
float16 : 5.15
Single-precision compute (GFLOPS)
float : 25.12
float2 : 47.47
float4 : 46.81
float8 : 41.68
float16 : 46.45
Half-precision compute (GFLOPS)
half : 23.11
half2 : 50.10
half4 : 98.10
half8 : 93.34
half16 : 88.61
Double-precision compute (GFLOPS)
double : 5.85
double2 : 3.27
double4 : 20.97
double8 : 20.65
double16 : 20.40
Integer compute (GIOPS)
int : 22.67
int2 : 49.85
int4 : 47.96
int8 : 34.30
int16 : 47.78
Transfer bandwidth (GBPS)
enqueueWriteBuffer : 1.24
enqueueReadBuffer : 1.31
enqueueMapBuffer(for read) : 5.22
memcpy from mapped ptr : 2.56
enqueueUnmap(after write) : 5.32
memcpy to mapped ptr : 2.73
Kernel launch latency : 307.37 us
Troubleshooting
- Post your issue at the forum: https://forum.radxa.com/c/dev