The GNU Compiler Collection has an NVIDIA PTX target, supporting both offload modes for OpenMP and OpenACC on NVIDIA GPUs. It also supports a quite… Read More »GCC’s NVPTX target

Side note: OpenCL support doesn’t ship at all for NVIDIA Jetson platforms. As such, CUDA is the only available low-level GPGPU API there. NVIDIA’s embedded… Read More »CUDA on NVIDIA Jetson – just not the same

Android is the most popular operating system in the world, rolled out on billions of devices. What is the level of flexibility allowed across across… Read More »Standards on Arm platforms for OS support – Android

AIR is the LLVM IR dialect used by Metal. AIR shader binaries are guaranteed to run unmodified on a future operating system release. However, Apple… Read More »OpenCL on Metal #3: Metal compilation

For April Fools day, an alternate approach is discussed for the OpenCL on Metal project. Writing an OpenCL implementation from scratch would take a long… Read More »OpenCL on Metal #2 – what if clvk works?

Habana is an Intel company that designs machine learning accelerators for both training and inference. Their kernel driver is available upstream in the Linux kernel.… Read More »Open-source drivers according to Habana

OpenCL today is considered as deprecated on macOS, let’s see if we can get OpenCL kernels compiled to AIR to run on Metal. Let’s start… Read More »OpenCL on Metal #1 – get_global_id()

For iOS simulators, dyld_sim is used as the dynamic linker. How small can the list of dependencies be to run a “hello world” in such… Read More »What is the minimum number of libraries to run a “hello world” on iOS Simulator?

On recent driver revisions for Linux x86_64 (this does not apply to AArch64 NVIDIA drivers at this time), NVIDIA shipped firmware for a new coprocessor,… Read More »NVIDIA GSP Firmware