I'm currently working in my master thesis (Distributed Multi-GPU k-Nearest-Neighbor algorithm using Locality Sensitive Hashing and SYCL). I've already tested my implementation with ComputeCpp, hipSYCL and oneAPI.
I'm currently working on a table containing possible hardware targets and how each implementation supports those targets.
For example for ComputeCpp:
Intel CPUs: OpenCL (SSE4.1 reuqired)
AMD CPUs: OpenCL
NVIDIA GPUs: CUDA + PTX (experimental)
AMD GPUs: OpenCL + SPIR
Intel GPUs: OpenCL
Intel FPGAs: N/A
Xilinx FPGAs: N/A
Now I'm not sure about the oneAPI support and implementation for each of these hardware targets:
Intel CPUs: OpenCL
AMD CPUs: OpenCL (are AMD CPUs supported?)
NVIDIA GPUs: CUDA + PTX
AMD GPUs: are AMD GPUs supported?
Intel GPUs: OpenCL + SPIR
Intel FPGAs: OpenCL + SPIR
Xilinx FPGAs: are other FPGAs supported too?
Sorry if this is the wrong way to ask such a question.
I would be very happy to receive an answer.
Thanks for reaching out to us.
As of now, Intel OneAPI targets only Intel Specific Processors and Accelerators. You can find all the supported devices here:
However, there is a Codeplay contribution to dpcpp through which you can target NVIDIA GPU's(only). For more info on how to use it you can refer to this below article:
We also have a detailed OneAPI specification which you can refer: https://spec.oneapi.com/versions/latest/index.html
Please reach out to us if you face any issue or if you have any more questions.
We are closing this thread considering your query has been answered.
Please raise a new thread for any further queries.
Any further interaction in this thread will be considered community only