Device and DeviceType

PyTorch provides device abstractions for writing code that works across CPU, CUDA, and other backends.

Device

{doxygenstruct}

:members:
:undoc-members:

Example:

cpp

c10::Device cpu_device(c10::kCPU);
c10::Device cuda_device(c10::kCUDA, 0);  // CUDA device 0

if (cuda_device.is_cuda()) {
    std::cout << "Using CUDA device " << cuda_device.index() << std::endl;
}

DeviceType

{cpp:enum-class}


Enumeration of supported device types.

{cpp:enumerator}


CPU device.

{cpp:enumerator}


NVIDIA CUDA GPU.

{cpp:enumerator}


AMD HIP GPU.

{cpp:enumerator}


XLA / TPU.

{cpp:enumerator}


Vulkan GPU.

{cpp:enumerator}


Apple Metal GPU.

{cpp:enumerator}


Intel XPU GPU.

{cpp:enumerator}


Apple Metal Performance Shaders.

{cpp:enumerator}


Meta tensors (shape only, no data).

{cpp:enumerator}


Habana HPU.

{cpp:enumerator}


Lazy tensors.

{cpp:enumerator}


Graphcore IPU.

{cpp:enumerator}


Meta training and inference accelerator.

{cpp:enumerator}


Custom backend registered via `c10::register_privateuse1_backend()`.

Convenience constants:

c10::kCPU, c10::kCUDA, c10::kHIP
c10::kXLA, c10::kVulkan, c10::kMetal
c10::kXPU, c10::kMPS, c10::kMeta
c10::kHPU, c10::kLazy, c10::kIPU, c10::kMTIA
c10::kPrivateUse1