GPU Flow

GPUs are special-purpose compute devices that can be used to offload a compute intensive portion of your application. GPUs usually consists of many smaller cores and are therefore known for massive throughput. There are some tasks better suited to a CPU and others that may be better suited to a GPU.

Note

In addition to Intel® processors listed in the System Requirements, AMD* and NVIDIA* GPUs may also be targeted (Linux only):

Tip

Unsure whether your workload fits best on CPU, GPU, or FPGA? Compare the benefits of CPUs, GPUs, and FPGAs for different oneAPI compute workloads.