Intel's New Acceleration Architecture Is Quick To Advance
It was only four days ago that Intel introduced the Sandy Bridge New Acceleration architecture, which brought tremendous speed improvements to their open-source Linux driver stack. Phoronix benchmarks are still forthcoming, but in many workloads the improvements are absolutely incredible, not only for the latest Sandy Bridge hardware but all generations of supported Intel integrated graphics.
Obviously, the first-cut "SNA" acceleration architecture is still a work in progress. However, every day since the initial push has continued to see new improvements and fixes.
On this page is a list of commits to the xf86-video-intel DDX that are in regards to the the "Sandy Bridge New Acceleration." The biggest work since the push of this new acceleration model took place less than 24 hours ago and it adds Zaphod mode support under the SNA architecture. Zaphod mode is a way of creating an X Server with multiple screens from a single device and is similar to the implementation found with the other open-source DDX drivers. Zaphod mode though is not the supported way of dealing with multiple screens, but RandR 1.4 with its per-CRTC-pixmaps is the better solution when available.
Benchmarks of this new acceleration code are coming soon.
Obviously, the first-cut "SNA" acceleration architecture is still a work in progress. However, every day since the initial push has continued to see new improvements and fixes.
On this page is a list of commits to the xf86-video-intel DDX that are in regards to the the "Sandy Bridge New Acceleration." The biggest work since the push of this new acceleration model took place less than 24 hours ago and it adds Zaphod mode support under the SNA architecture. Zaphod mode is a way of creating an X Server with multiple screens from a single device and is similar to the implementation found with the other open-source DDX drivers. Zaphod mode though is not the supported way of dealing with multiple screens, but RandR 1.4 with its per-CRTC-pixmaps is the better solution when available.
Benchmarks of this new acceleration code are coming soon.
4 Comments