PolarDB:Enable GPU-accelerated computing

GPU parallel acceleration in a database operates at the object level. It converts the objects of a single field into a model that is suitable for parallel computing, and leverages the massive number of GPU cores to perform parallel computation.

GanosBase automatically enables GPU-accelerated computing when it detects a GPU device. No additional parameter settings are required on the implemented functions, and the accelerated computing is transparent to users. At the same time, GanosBase provides session-level control permissions, allowing users to decide whether to enable GPU-accelerated computing.

1. Verify whether the current environment has a GPU device.

   1. Create the ganos_raster extension create extension ganos_raster cascade.

   2. Execute the SQL statement select st_checkgpu().

      • If a GPU device is detected in the current environment, the GPU device information is returned.

        rasterdb=# select st_checkgpu();
                            st_checkgpu
        ------------------------------------------------------------------------------------------------------------------------------
         [GPU(0) prop]multiProcessorCount=20; sharedMemPerBlock=49152; totalGlobalMem=-608239616; maxThreadsPerBlock=1024; maxThreadsPerMultiProcessor=2048; cudaThreadGetLimit=1024 .
        (1 row)

      • If no GPU device is detected in the current environment, an error message is returned.

        rasterdb=# select st_checkgpu();
                           st_checkgpu
        ------------------------------------------------------------------------------------------------------------------------------------------
        There is not gpu device on current environment, cuda_errorcode=35, errormsg=CUDA driver version is insufficient for CUDA runtime version.
        (1 row)

      Note

      GPU-accelerated computing can be enabled only in an environment with a GPU device.

2. Enable and disable the session-level GPU usage status.

   • In an environment with a GPU device, GanosBase enables GPU-accelerated computing by default. If you want to disable GPU-accelerated computing and use the original CPU computing mode, execute set ganos.raster.use_cuda=off in the session:

     rasterdb=# set ganos.raster.use_cuda=off;
     SET
     rasterdb=# show ganos.raster.use_cuda;
      ganos.raster.use_cuda
     -----------------------
      off
     (1 row)

   • If you need to re-enable GPU-accelerated computing, execute set ganos.raster.use_cuda=on:

     rasterdb=# set ganos.raster.use_cuda=on;
     SET
     rasterdb=# show ganos.raster.use_cuda;
      ganos.raster.use_cuda
     -----------------------
      on
     (1 row)

3. The modules that implement GPU-accelerated computing in GanosBase.

   Note

   Currently, GPU-accelerated computing is implemented only in the Raster module of GanosBase. Support for the Trajectory and Geometry modules will be added in the future.