1. Computers
  2. Display Drivers
  3. Graphics Cards
  4. Memory
  5. Motherboards
  6. Processors
  7. Software
  8. Storage
  9. Operating Systems


Facebook RSS Twitter Twitter Google Plus


Phoronix Test Suite

OpenBenchmarking.org

AMD Llano Graphics Battle: Gallium3D vs. Catalyst

Michael Larabel

Published on 26 August 2011
Written by Michael Larabel
Page 1 of 7 - 4 Comments

A month ago we looked at the Radeon HD 6550D graphics performance from the AMD Fusion A8-3850 (a new "Llano" APU) under Linux when using the Catalyst driver. However, bugs at the time had barred a comparison of the Llano graphics under Linux with the open-source Mesa/Gallium3D driver. Fortunately, we now have a working open-source Radeon driver configuration to deliver these comparative AMD Llano Linux OpenGL benchmarks.

As reported back in July, there were display issues when attempting to use the open-source Radeon DRM driver. When attempting to mode-set the display with any of the outputs, the display would become garbled and simply not work out right. Earlier this month the Llano support was still fairly broken, but AMD's Alex Deucher on his production board was at least able to get the DVI and HDMI outputs working while VGA would outright fail.

For Llano users having such issues with the very latest Linux kernel code, Alex recommends trying a combination of DVI and HDMI. Toggling dual-link DVI support from the BIOS may also help out, he says. This week when finally pulling in all of the latest code (the Linux kernel, xf86-video-ati DDX, and Mesa from Git master as of 24 August), and going through each display configuration on two Llano motherboards, a working configuration was found.

What was attempted first was using the Gigabyte GA-A75M-UD2H motherboard with the A8-3850. Dual-link DVI was enabled from the BIOS and using the DVI connection a single-link display (Dell LCD; 1920 x 1080) and dual-link display (Samsung 30-inch LCD; 2560 x 1600) were tried.

For a short time (circa 60 seconds) it would be possible to use the system after mode-setting took place, but then eventually there would be a hard lock-up and the frame-buffer became corrupted. There was at least working 3D acceleration for the short time the system worked.

Latest Articles & Reviews
  1. Khronos Group Announces Vulkan, OpenCL 2.1, SPIR-V
  2. Samsung 850 EVO SSD Linux Benchmarks
  3. Kubuntu 15.04 Is Turning Out Quite Nice, Good Way To Try Out The Latest KDE
  4. 5-Way Linux Distribution Comparison On The Core i3 NUC
  5. OCZ ARC 100 Linux SSD Benchmarks
  6. Lenovo ThinkPad X1 Carbon Works Great As A Linux Ultrabook
Latest Linux News
  1. Ubuntu Cloud Switches Over To Using Systemd By Default
  2. System76 Introduces The Broadwell-Powered, Ubuntu-Loaded "Meerkat"
  3. Xfce 4.12 Might Make It For Fedora 22
  4. Pictures Of The Near Production Ready Ubuntu Tablet
  5. OpenVG Support Stripped From Gallium3D
  6. Mozilla Is Getting Excited About WebGL 2
  7. ZOTAC Announces Steam Machine SN970, Powered By SteamOS
  8. Allwinner Continues Violating The LGPL
  9. AMD Gets Into VR With LiquidVR Technology
  10. NVIDIA Announces Shield Set-Top Gaming Box
Most Viewed News This Week
  1. More Proof That Allwinner Is Violating The GPL
  2. The Tremendous Features Of Fedora 22
  3. Krita 2.9 Released, Their Biggest Release Ever
  4. A Single UEFI Executable With The Linux Kernel, Initrd & Command Line
  5. Confirmed: Vulkan Is The Next-Gen Graphics API
  6. Xfce 4.12 Released After Nearly Three Years Of Work
  7. LLVM 3.6 & Clang 3.6 Deliver More Features, Complete C++14 Support
  8. 8cc: A Small C11 Compiler