CodePlay creates SyCl Academy
Kronos Group release documentation for OpenCl C Extensions and OpenCl C++
New Clang 10 supports OpenCl C++ and OpenCl Extensions

Efficient GPU programming with C++
Rocm 2.9 is out

Kronos Group opened a new github channel for SYCL extensions/proposals/materials
AMD annunces Rocm 3.0 with CUDA conversion with HIP, Tensorflow and Pytorch support, etc.
First beta of Intel One-Api, based on SYCL.
Codeplay released a new version of ComputeCpp

AMD annunces Rocm 3.0 with CUDA conversion with HIP, Tensorflow and Pytorch support, etc.

Rocm 3.0 is [url=https://github.com/RadeonOpenCompute/ROCm]here[/here]

Pocl 1.4 with LLVM 9 support
Intel OCLOC now support OpenCL/Spir-V for multiple device
PyOpenCl 2019.1.2

Intel released new SDK for OpenCL

AMD released Rocm 3.3 with ONNX enhancements
PoCL 1.5 with LLVM 10 support
Microsoft is working to port OpenCl mapping layers on DirectX12

Kronos releases provisional OpenCl 3.0 specifications

C++ and Sycl for heterogenous programming

It is difficult to support OpenCL for graphics cards that do not support OpenCL,
which requires the graphics card itself to open a certain degree of support for the underlying hardware.
But it's not entirely impossible.
If you need a smart IC chip to participate in the production, you can contact us
URL:https://www.fpgamall.com/

It is difficult to support OpenCL for graphics cards that do not support OpenCL

Amd, Nvidia and Intel support OpenCl since 2009 (on professional and consumer gpu).
Arm gpu Mali supports OpenCl since 2012.
So, it's NOT an hardware problem.

Latest top level Nvidia Gpu produces 30 Tflops in single precision.
Latest top level Amd Gpu produces 10 Tflops in single precision.

Latest top level Intel cpu (I9-10980XE) produces 1 Tflops in single precision.

ROCm 3.7 has Open-Source OpenCL Image Module included
Oak-Ridge supports GCC improvements on GPU offloading

Latest release of oneAPI

I'd expect both Ralph@home and Rosetta@home to use only OpenCL compilers that can run under Linux but produce code to run under either Linux or Windows.

Also, remember that the last time they tried, the GPU code would run only at approximately at the same speed as the CPU code, rather than ten times the speed of the CPU code which is usually the minimum accepted for BOINC projects.

Also, remember that the last time they tried, the GPU code would run only at approximately at the same speed as the CPU code, rather than ten times the speed of the CPU code which is usually the minimum accepted for BOINC projects.

I know, i know. But the "last time they tried" was, if i'm not wrong, over 7 years ago.
7 years in gpgpu is like an aeon, not only for hw (considering, for example, the computational power of recent RTX3090 and the incoming AMD top gpu), but also for sw, libraries, frameworks, etc (you can think, for example, to new SYCL language).

AMD releases AOMP 11.9

Rocm 3.8 with a lot of new features

Intel, with the Heidelberg University, is working on port oneAPI/DPC++ on AMD Gpu thanks to HypSycl
Codeplay is working on port oneAPI/DPC++ on Nvidia Gpu thanks to SYCL