While Intel's OSTC (Portland) team is busy at work on Intel Ivy Bridge Linux graphics support for this next-generation hardware due out by year's end, the same team doesn't play with Intel's Poulsbo or other graphics IP that isn't an in-house Intel creation and part of their open-source driver. It seems, however, that Alan Cox is personally working on early "Cedar Trail" support for the open-source GMA500 driver.
Last week I mentioned that it looked like Google was interested in Gallium3D for Chromium OS. That has turned out to be the case and the flow of changes to the community Intel Gallium3D driver has only continued to increase.
The open-source developers working on the drivers for AMD/ATI Radeon and NVIDIA (via the Nouveau project) graphics hardware have tossed all their weight behind the Gallium3D driver architecture. The Gallium3D drivers have surpassed the "classic" Mesa DRI drivers in terms of capabilities, performance, and stability. The only strong holdout to Gallium3D has been Intel since they aren't convinced that it's the appropriate choice and they aren't interested in overhauling their Linux driver stack once more with the large upfront investment that's required in rewriting their user-space 3D driver in moving from classic Mesa to Gallium3D.
Intel's current-generation "Sandy Bridge" processors continue to sell incredibly well and perform phenomenally relative to AMD's current offerings and Intel's previous-generation hardware. Under Linux, the Sandy Bridge support is now excellent if pulling in the latest components (namely the Linux kernel, xf86-video-intel, and Mesa) and only continues to be improved over time with advancements like their new driver acceleration architecture. By year's end, Intel is expected to launch their "Ivy Bridge" processors as the successor to Sandy Bridge. Intel is already preparing the Ivy Bridge Linux support code.
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.
While out celebrating the 7th birthday of Phoronix, Intel pushed out a new acceleration architecture for their open-source Linux driver. This new acceleration architecture is called "SNA" for "SandyBridge's New Acceleration", and it brings incredible results not only for Sandy Bridge, but for previous generations of Intel graphics as well. The results provided by Intel are absolutely stunning.
Beyond the demo of MeeGo Tablet UX running on Wayland, next week at the MeeGo Conference in San Francisco are a variety of other interesting sessions.
The MeeGo conference is running from tomorrow through Wednesday in San Francisco. This is the first conference for the Moblin-Maemo-mix since Nokia parted ways to team up with Microsoft and ship Windows Phone 7 on their future devices, but there's interesting work still going on in the MeeGo world. In particular, of interest to many Phoronix readers will be the fact that it sounds like the adoption of the Wayland Display Server is going quite well within the MeeGo world. It appears that there's already an experimental version of MeeGo Tablet UX working atop Wayland.
Prior to being spun into MeeGo, Intel's Moblin Linux distribution was one of the fastest-booting Linux distributions. Moblin worked phenomenally for Intel Atom netbooks with a great user-interface and was very quick at starting up. With the release this week of MeeGo 1.2, we have some Bootchart numbers for MeeGo 1.2 Netbook UX compared to Fedora 14 and Ubuntu 11.04.
Intel, now without the support of Nokia, has released their MeeGo 1.2 Linux-based operating system. The MeeGo 1.2 core release targets Intel Atom and ARMv7 architectures and is meant to serve as a baseline for device vendors and software developers to base their work upon.
This morning I wrote about the troublesome experience of Intel Sandy Bridge graphics under Ubuntu 11.04 as the packages found in the Natty repository are outdated and contain only the initial "SNB" support. In the mainline upstream code, Sandy Bridge is supported much better, offers faster performance, and possesses other new features (e.g. VA-API encode), except in the past week the Intel SNB Linux code temporarily broke hard.
The Intel G45 chipset was released in the summer of 2008, but only this week is it now possible to take advantage of VA-API video playback acceleration for this Intel integrated graphics processor.
Intel's Keith Packard has issued a pull request to David Airlie for their drm-intel-next branch for the Linux 2.6.40 kernel. The drm-nouveau-next was just pulled as well.
Intel has released an OpenCL Software Development Kit that provides support under Linux.
While some of Intel's Atom processors use PowerVR graphics (the notorious Poulsbo and now Moorestown), the vast majority of the Atom CPUs on the market take advantage of Intel's own graphics technologies developed in-house. This allows Intel to provide Linux support via their mainline open-source driver and overall the support is quite good. However, it seems with the next-generation Atom CPUs, this will change.
Ivy Bridge is Intel's next-generation processor to succeed Sandy Bridge by the end of this calendar year. At the end of April there was the release of open-source Ivy Bridge for the DRM/KMS driver in the Linux kernel so that the support can land in advance of the hardware's availability. Just moments ago Intel has now pushed out the open-source DDX (X.Org driver) support for Ivy Bridge as well.
Beyond working on initial Ivy Bridge support and continuing to fortify the Sandy Bridge support, Intel's OSTC team is still working on improving the GLSL compiler for Mesa.
Just as I had said two weeks ago: expect Intel "Ivy Bridge" open-source Linux support to be pushed to the Linux 2.6.40 kernel. This code was pushed today in preparation for the Linux 2.6.40 kernel. It will land in a similar manner to Valve's Source/Steam native Linux support coming soon, early AMD Bulldozer benchmarks, ATI/AMD's "Orka" Linux driver support in the past, etc. Here's Intel's Ivy Bridge code that enables next-generation hardware support.
In continuation of Mobile Users Beware: Linux Has Major Power Regression, what's causing this regression in the mainline Linux kernel between Linux 2.6.37 and 2.6.38 (and still outstanding in 2.6.39) is still a mystery. To some surprise, even when disabling power management on one of the ThinkPads, the problem is still present.
With xf86-video-intel 2.15 now having been released, Intel has just issued the 2011Q1 Linux graphics package as their quarterly open-source update.
Intel's quarterly driver package for Q1'2011 is about to be released and comprised as part of that is the xf86-video-intel DDX driver. Chris Wilson has just announced the release of xf86-video-intel 2.15.0 for this release.
Long ago there was a promise by Intel's Linux developers that they would deliver VA-API video acceleration support to G45 IGPs in the second quarter of 2010. This didn't happen even while support for VA-API on Clarkdale/Arrandale arrived in early 2010. There's also now VA-API support for Sandy Bridge and even video encoding support. But the old G45 hardware hasn't received much love. It now appears though that Intel might release VA-API for G45 in the coming months.
Yesterday I mentioned the Intel Sandy Bridge video encode support should be here soon for Linux users and that it's in a working state after an Intel employee accidentally posted a test report to the driver's public mailing list. It turns out the code is actually available right now.
Since Intel began shipping Intel Core i3/i5/i7 "Sandy Bridge" CPUs back in January, H.264 VA-API video (decode) playback acceleration has been available if using all of the proper open-source Linux packages needed for support. The video encode support though wasn't there on time nor was there much in the way of official communication in what to expect or when it would come, but the word received at Phoronix was that it would be a Q1 target. It looks like the code for VA-API video encode on Sandy Bridge is nearly ready.
Chris Wilson has just tagged the second release candidate of xf86-video-intel 2.15.0.
Near the beginning of this month I talked about an important Sandy Bridge performance fix landing in Mesa that with 13 lines of changed code resulted in a huge performance improvement for those using the new integrated graphics found on the Sandy Bridge CPUs. This performance boost was quite dramatic and made the open-source Intel Linux driver comparable to Intel's closed-source Windows driver, but the performance tuning is not done yet. There's DRM patches arriving this morning that squeeze more performance out of Intel Sandy Bridge graphics under Linux.
With it nearing the end of the quarter, Intel's OSTC team working on their Linux graphics driver stack is readying their quarterly driver update. Along with the Linux 2.6.38 kernel and Mesa 7.11 as some of the key components to make up this quarter's Linux package, the xf86-video-intel 2.15.0 X.Org driver will also be released. In preparing for this milestone, Chris Wilson has released the first development snapshot of this DDX driver.
While there is Intel VA-API video acceleration support under Linux for Clarkdale/Arrandale hardware and the newest Sandy Bridge CPUs (assuming you are running the very latest code), there is no video playback acceleration support for the Intel G45 / GMA 4500M HD hardware. It was previously promised by Intel engineers with a target delivery date of Q2'2010, but that has long since passed without any further information from Intel.
The Mesa code-base now has patches for supporting the new MESA_multithread_makecurrent GLX extension. This extension was originally proposed in 2009 at the Cairo and GStreamer Hackfest but more than a year later it's now only materializing within Mesa and first within Intel's driver.
Following last week's announcement of Microsoft and Nokia hooking up over Windows Phone 7 for Nokia's future products, left MeeGo in a bad position (along with Qt), but Intel has now commented today that MeeGo is still going strong. Miguel de Icaza has also provided his commentary praising this move by Nokia, yes, really.
997 Intel news articles published on Phoronix.