• Optimize LC0 - Leela Chess Zero - for AMD GPUs

    Heyho AMD community,   we are all aware about the neural network hype on gpus, and most have noticed that Nvidia has simply the forehand with their cuDNN framework.   Personally I am convinced that AMD mak...
    created by smato2018
  • Newbie introduction and seeking guidance.

    Hello my name is Ernst. I have a passion for binary level data encoding.  I work on my workstation running a AMD 9590 and recently I made the decision to upgrade my NVIDIA GPU to twin Radeon wx 5100s pros'. I am...
    last modified by ernst0
  • Need tips to hide memory latency - 20x speed-loss when writing to memory

    I have an OpenCL code that does Monte Carlo photon transport simulations in a voxelated space (https://github.com/fangq/mcxcl). The code involves simulating a large number of random photon trajectories, each in a thre...
    last modified by FangQ
  • Running OpenCL Work Groups with >256 Elements

    Hi all,   I am currently re-writing some OpenCL code of mine and would like to split the work of the group to more waves in order to have more waves in flight. The code is a OpenCL 1.2 code (because it needs to ...
    last modified by lolliedieb
  • Opengl interop - chosen wrong device but works

    Hello. I have quite old hardware in my laptop with switchable graphics - Intel HD 4000 and Radeon HD 7670M. Switchable graphics works but behaves unexpectedly. I have following code to choose opencl device for textu...
    last modified by omega_doom
  • OpenCL compilation hangs forever

    Hi all,   I am trying to compile this project for an AMD GPU: GitHub - webmaster128/lisk-vanity: A tool to generate short Lisk addresses with GPU support   The c.l files are in lisk-vanity/src/opencl at m...
    last modified by webmaster128
  • Kernel runs slower for local workgroup size greater than 64

    Hi bros, I'm a CS undergraduate student and I recently wrote a GPU path tracer using OpenCL. If you don't know what path tracing it's basically a method to generate photorealistic images by shooting rays through every...
    last modified by gallickgunner
  • Processing two buffers using an out of order queue

    I have a PCI data acquisition card that supports P2P. It will be capturing records one after the other at a very rapid rate, and the plan is to write each record to the GPU using DirectGMA, where a kernel will process...
    last modified by andyste1
  • Realtime raytracing with opencl

    Is it possible to achieve realtime raytracing like RTX with opencl?
    created by cyseal
  • List of neural network/machine learning/GPU computing apps that support OpenCL acceleration on AMD Fx HW?

    Hi, I have a few questions. I hope you can help me.   I am trying to learn neural nets/ML on my older, Fx based hardware.   I very much prefer the openCL development model. As discussed elsewhere, people ...
    last modified by devlista
  • Modern GPU: book-length tutorial and open-source GPGPU library

    Fastest GPU radix sort and scan-centric tutorial. Hi all. I've been putting together a big book-length online GPU computing tutorial. It's at http://www.moderngpu.com/ The content is very scan/reduction-centri...
    last modified by BarnacleJunior
  • Please add new extension for refined reduce in wavefront

    Hi,   According to https://gpuopen.com/amd-gcn-assembly-cross-lane-operations/  the hardware is able to do refined reduce operations.   By 'refined', I have in mind doing an add/min/max among neighbor...
    last modified by mannerov
  • Strange behavior of a kernel, need fresh ideas

    I lost two days debugging or better to say tried to debug my kernel. Basically the kernel looks like this (part of dagger-hashimoto initialization):   1. copy from global to private 2. do private 3. copy from ...
    last modified by kbala
  • Missing OpenCL CPU support under Windows

    On my system the (i think) most recent version of the AMD drivers (18.8.1, Windows 10 x64) no longer returns the CPU (FX-8350) as a valid OpenCL device.  Is this intended behavior or just a bug in my specific ins...
    last modified by pangea
  • How to tune the performance of ROCm(llvm) compiler?

    I modified llvm (roc-1.6.x) a bit to generate a code that can run on AMDGPU pro dirver. It can run but the performance is over 10% slower than AMDGPU's online compiler, for the same opencl code.  I wonder if ther...
    last modified by fancyix
  • Disappointing opencl half-precision performance on vega - any advice?

    I bought a Vega 64 recently. From the specs, it has 23 TFLOPs fp16 throughput compared to 12 TFLOP fp32. so I converted portion of my Monte Carlo code to half, expecting to gain some noticeable speed up. Disappointing...
    last modified by FangQ
  • How to controll opencl kernel configuration of assembly code generated by llvm clang AMDGPU backend?

    llvm clang can compile opencl file into assembly. A common format is hsa. There are certain configurations in hsa assembly file, such as enable_sgpr_dispatch_ptr and enable_sgpr_queue_ptr. When I compile my opencl fil...
    last modified by fancyix
  • How to use llvm to offline compile .cl file into binary that can run with amdgpu pro driver?

    Now I am trying to build OpenCL kernel binary with llvm. I successfully compiled .cl into assembly, but cannot figure out a way to compile that format of assembly into binary that can run with AMDGPU pro driver. That ...
    last modified by fancyix
  • Can OpenCL build program from .cl source code together with a pre-built binary kernel?

    My program has several kernels. I'd like to use offline compiler to compile one kernel into binary. So how can I build my program using other kernels and that one pre-built kernel binary?
    created by fancyix
  • How to compile .cl file that contains inline assembly for GCN cards?

    There are some examples of inline assembly inside .cl file: LLVM-AMDGPU-Assembler-Extra/s_memrealtime_inline.cl at master · ROCm-Developer-Tools/LLVM-AMDGPU-Assembler-Extra · GitHu… gatelessgate/equ...
    last modified by fancyix