Clik here to view.
HSA - Sample code - Creating the queue failed
Hello. I am getting my machine setup with HSA for a project. I managed to follow the instruction installation steps at the HSA's Github page. I am now trying to run the sample program vector_copy. It...
View ArticleClik here to view.
AMD-GPU PRO and ROCm (Ubuntu 16.04 x64)
Is it possible to get both ROCm and AMD-GPU PRO driver stacks to run at the same time?
View ArticleClik here to view.
ROCm with RX480 "Ellesmere"
Dear All, I'd like to ask you that is can I run ROCm platform on the latest RX480 hardware? Has anyone tried it already?Also I'm not really familiar with HCC compiler, but it is possible to perform...
View ArticleClik here to view.
ROCm windows support
Hey guys, do you know if there is an upcoming Windows support for the ROCm platform?
View ArticleClik here to view.
Strip .brig section from ELF binary
When I convert my OpenCL kernel to binary, I can set flags to remove .source, .llvmir and .amdil sections from ELF.Is there a way of removing the .brig section as well?
View ArticleClik here to view.
Can i use GCN Assembly kernels in the same why as HSAIL kernels?
TL;DR:Can i use Kernels written in GCN Assembly with HSA? I like to have thefeature of persistent kernels and feed them work if needed. Hi, im planning a project and for preperation i was reading a lot...
View ArticleAMD HSA - huma with extra dedicated GPU?
Hi,i got an AMD A10-7850K and wanted to test and play withGitHub - HSAFoundation/CLOC: CL Offline Compiler : Compile OpenCL kernels to HSAIL But i also thought about getting a dedicated GPU to...
View ArticleBuild HIP on Nvidia platforms(use nvcc instead of hcc)
Hi all, I'm trying to build HIP for two of my boxes running Manjaro Linux. A GTX 970 is installed on one machine and the other has a GTX 780 Ti.But the building guild(HIP/INSTALL.md at master ·...
View ArticleHardware support (ryzen + GFX7)
Hi, The readme (Update README.md · RadeonOpenCompute/ROCK-Kernel-Driver@a75fb71 · GitHub ) and hardware support list ROCm, a New Era in Open GPU Computing do not seem to be aligned - readme states...
View ArticleKaveri CPU-iGPU connections
I have some questions regarding how the CPU and integrated GPU interact within the memory hierarchy. I highlighted the keyword items in question.1. Does the integrated GPU have a DMA controller ?...
View ArticleQestions of ROCm and the relation between it and HSA
Hello all,After doing some research online and reading the introduction of ROCm. I'm confused on what HSA (AMD HSA Runtime) and ROCm does. They seems to both provide the same thing(HSA APIs) and that...
View ArticleClik here to view.
Explanation of terms from chapter 9 of GCN ISA manual
Hello, I am not allowed to post in the HSA section (or is there a GCN/ROCm section?), hopefully someone knowledgeable will find the question here. Chapter 9 of GCN ISA Manual (the version from August...
View ArticleROCm documentation feedback
In case it is helpful to ROCm doc editors, attached are some questions, formatting problems, miswordings, and typos/misspellngs I encountered while reading the first pages (Welcome, Release Notes,...
View ArticleClik here to view.
HSA - Sample code - Creating the queue failed
Hello. I am getting my machine setup with HSA for a project. I managed to follow the instruction installation steps at the HSA's Github page. I am now trying to run the sample program vector_copy. It...
View ArticleGood way to share carry flag cross lane?
I need to add two ulong.The first half (low 32 bits) of each number is stored in lanes with even id, and the second half (high 32 bits) is stored in the lane next to it (next odd id).So one lane adds...
View ArticleGet element from an array in a paticular lane
Say I have 8 threads, I want to get array tsum[laneID%8] from lane 0/8 for thread of laneID.One possible way is to use 16 move dpp instructions with row shift, like this:__asm ( \ "s_nop 1\n" \...
View ArticleGPUOpen Anaconda links broken
On the pagehttps://gpuopen.com/compute-product/continuum-analytics-anaconda-with-numba/ The document link is brokenhttp://numba.pydata.org/numba-doc/latest/hsa/index.html (404 Page not found) Maybe it...
View ArticleClik here to view.
"Radeon Open Compute Platform (ROCm)" link is broken
On the community HSA overview page, the "Radeon Open Compute Platform (ROCm)" link is broken.It leads tohttps://radeonopencompute.github.io/ (404 Site not found) but maybe it should lead...
View ArticleClik here to view.
Tensorflow with AMD GPU
I just bought a new Desktop with Ryzen 5 CPU and an AMD GPU to learn GPU programming. I am also interested in learning Tensorflow for deep neural networks. After a few days of fiddling with tensorflow...
View ArticleClik here to view.
ROCm readthedocs column width, table wrapping
The ROCm documentation on the readthedocs io site is written in reStructuredText (.rst).It is compiled and uses the readthedocs sitewide theme.css. The theme.css is designed to be responsive to varying...
View Article