Search notes:

GPU

A GPU is specialized for compute-intensive, highly parallel computations.
In order to achieve this goal, compared to a CPU, more transistors are devoted to data processing rather than data caching and flow control.

Single Instruction, Multiple Threads (SIMT)

Nvidia introduced SIMT in the Tesla GPU microarchitecture with the G80 chip.
ATI Technologies, now AMD, released a competing product slightly later on May 14, 2007 (the TeraScale 1-based "R600" GPU chip).

Determine GPU type

Linux

$ lspci | grep -i vga

Windows

In Windows (11?), the GPU type can be determined with
  • Task manager under Performance
  • Settings -> System -> Related settings -> Advanced display (Display information)

iGPU vs dGPU

iGPU = integrated GPU (integrated into the CPU, that is)
dGPU = dedicated (or discrete?) GPU

See also

torch.nn.Module.cuda()
tensorflow.config.list_physical_devices('GPU')
torch.cuda.device_count()
The --gpus option of docker run.
TPU

Links

nvtop is a (h)top like task monitor for AMD, Intel and NVIDIA GPUs.

Index