Programmatically limiting GPU access

[cburton]

Is it possible to limit access to the GPU programmatically, either making it unavailable or only available at a specified limit?

[AastaLLL]

Hi,

Jetson doesn’t have a mechanism to control GPU resource limit.
But you can make it unavailable with the below global variable.

$ ./deviceQuery
./deviceQuery Starting...

 CUDA Device Query (Runtime API) version (CUDART static linking)

Detected 1 CUDA Capable device(s)

Device 0: "Orin"
  CUDA Driver Version / Runtime Version          12.5 / 12.5
  CUDA Capability Major/Minor version number:    8.7
  Total amount of global memory:                 62841 MBytes (65893945344 bytes)
  (016) Multiprocessors, (128) CUDA Cores/MP:    2048 CUDA Cores
  GPU Max Clock rate:                            1300 MHz (1.30 GHz)
  Memory Clock rate:                             1300 Mhz
  Memory Bus Width:                              256-bit
  L2 Cache Size:                                 4194304 bytes
  Maximum Texture Dimension Size (x,y,z)         1D=(131072), 2D=(131072, 65536), 3D=(16384, 16384, 16384)
  Maximum Layered 1D Texture Size, (num) layers  1D=(32768), 2048 layers
  Maximum Layered 2D Texture Size, (num) layers  2D=(32768, 32768), 2048 layers
  Total amount of constant memory:               65536 bytes
  Total amount of shared memory per block:       49152 bytes
  Total shared memory per multiprocessor:        167936 bytes
  Total number of registers available per block: 65536
  Warp size:                                     32
  Maximum number of threads per multiprocessor:  1536
  Maximum number of threads per block:           1024
  Max dimension size of a thread block (x,y,z): (1024, 1024, 64)
  Max dimension size of a grid size    (x,y,z): (2147483647, 65535, 65535)
  Maximum memory pitch:                          2147483647 bytes
  Texture alignment:                             512 bytes
  Concurrent copy and kernel execution:          Yes with 2 copy engine(s)
  Run time limit on kernels:                     No
  Integrated GPU sharing Host Memory:            Yes
  Support host page-locked memory mapping:       Yes
  Alignment requirement for Surfaces:            Yes
  Device has ECC support:                        Disabled
  Device supports Unified Addressing (UVA):      Yes
  Device supports Managed Memory:                Yes
  Device supports Compute Preemption:            Yes
  Supports Cooperative Kernel Launch:            Yes
  Supports MultiDevice Co-op Kernel Launch:      Yes
  Device PCI Domain ID / Bus ID / location ID:   0 / 0 / 0
  Compute Mode:
     < Default (multiple host threads can use ::cudaSetDevice() with device simultaneously) >

deviceQuery, CUDA Driver = CUDART, CUDA Driver Version = 12.5, CUDA Runtime Version = 12.5, NumDevs = 1
Result = PASS

$ export CUDA_VISIBLE_DEVICES=1
$ ./deviceQuery
./deviceQuery Starting...

 CUDA Device Query (Runtime API) version (CUDART static linking)

cudaGetDeviceCount returned 100
-> no CUDA-capable device is detected
Result = FAIL

CUDA_VISIBLE_DEVICES is the variable that controls the visible GPU index.
Since Jetson only has one GPU (index=0), setting the variable with other values should meet your requirement.

Thanks.

[cburton]

Thank you. This implies that processes that had been launched prior to the environment variable having been set wouldn’t be aware of the setting?

[AastaLLL]

Hi,

Suppose the variable will take effect if the CUDA initialization is executed periodically.

Thanks.