|Introduction, Virtu Explained|
|Intel's Sandy Bridge processors have won acclaim for their hardware-accelerated video transcoding capabilities, dubbed Quick Sync, but there are significant limitations in the current iteration. LucidLogix, a company that's heretofore been known for its multi-GPU Hydra hardware, has developed a software-only solution named Virtu that it claims will solve one of the major issues that prevent enthusiasts from taking advantage of Sandy Bridge's transcoding performance, however.
The problem is this: At present, there's no way to easily use a discrete graphics card and access Sandy Bridge's transcoding engine. The only way to overcome this without using Virtu is to manually switch one's monitor cable from discrete GPU to integrated, reboot the machine, transcode, swap the cable back to the discrete GPU, and reboot again. Or use a multi-monitor setup.
Lucid's performance claims vis-à-vis ATI. We'll investigate them later in the article
When used in conjunction with an H57 motherboard, Virtu allows for both GPUs to function simultaneously, apart from a few quirks. Note that Virtu doesn't change anything with regard to Intel's enthusiast-oriented P67. Sandy Bridge's iGPU is automatically shut down when the chip is inserted into a P67 board and Intel claims it can't be changed.
Intel's upcoming enthusiast-oriented Z68 chipset will enable Quick Sync technology even if a discrete GPU is installed but Virtu, to the best of our knowledge, will still be required if a user wants both video transcoding via Quick Sync and discrete GPU performance.
Sandy Bridge's video processing hardware and its improvements over Clarkdale/Arrandale.
Virtu's high-level implementation is fairly simple; the program leverages some of the software assets Lucid has developed for Hydra. As the diagram indicates, Virtu is only compatible with DirectX. Games developed using OpenGL or WebGL would presumably run only on the Sandy Bridge iGPU.
Whenever a DirectX function call is made, Virtu evaluates the task and assigns it to either the dGPU or the iGPU. At this early stage the program's evaluation capabilities are still quite coarse. If the game in question has been flagged as a Virtu-compatible title, its rendering instructions are grabbed and tossed to the dGPU for execution. As the discrete card outputs its rendering results, Virtu snaps them up and copies them from the video card to the frame buffer associated with Sandy Bridge's iGPU.
Virtu is able to do this because the GPU core and frame buffer are decoupled on an abstract level. The GPU doesn't care what happens to a frame after it's been rendered, it just kicks the finished product out the door. Similarly, Sandy Bridge's frame buffer doesn't ask questions about where the data it's supposed to display just came from.
We suspect that the integrated PCI-Express controller on all Sandy Bridge processors is vital to Virtu's proper function. In the past, as with the X58, all PCIe links were attached to a northbridge or southbridge independent of the processor. The need to pull data through an extra 'hop' would've inevitably added latency to a product that depended on its absence.
Note that data earmarked for Sandy Bridge's native GPU isn't processed further and passes to the display buffer normally.
|Transcoding and Gaming Performance|
|The reason Virtu is considered such a big deal is simple: Sandy Bridge's iGPU transcodes video incredibly quickly; the speedup is similar to what we saw when Badaboom debuted. That program proved that leveraging Nvidia's CUDA technology could cut transcode times significantly compared to a standard CPU—without sacrificing image quality.
At present, typical users probably do precious little transcoding, but this could easily change in the not-too-distant future. The proliferation of tablets and media-rich smartphones will give consumers the ability to view content on a variety of devices. Scaling that content to fit both the TV and the iPhone requires transcoding and Sandy Bridge delivers in spades.
At present, only two programs support Intel's Quick Sync technology: Arcsoft Media Converter 7 and Cyberlink's MediaEspresso 6.5. We opted to focus on MediaEspresso, but have included performance results for the NV GTX 480 when using Badaboom as well. We transcoded a 720P rip of The Incredibles into an iPhone-compatible format and averaged the time-to-encode over three consecutive runs.
Our testbed consisted of an Intel Sandy Bridge 2600K processor (3.4GHz, Intel HD Graphics 3000), an Intel BOXDH55PJ H55 motherboard, 8GB of DDR3-1333 RAM in two 4GB sticks, and a 1TB WD Caviar Black. All tests were run multiple times and averaged.
We didn't expect to see MediaEspresso outpace Badaboom's CUDA-powered encode times by such a wide margin, but the gap was quite distinct. The 2600K only lags the GTX 480 in Badaboom by a small margin but we expect this is because we're encoding a low-resolution video. The gap between GPU and CPU transcoding performance will always vary depending on the type of encode being done and the various post-processing effects that are applied.
Quick Sync blows everything else away. At just over four minutes to encode a nearly two-hour movie it takes just 63 percent as long as its next-closest competitor, the GTX 480. The presence of the GTX 480 doesn't impact Quick Sync performance at all.
When it comes to Intel's transcode functionality, Lucid has things tightened up and good to go. Things are a bit less certain on the gaming side. We tested two games and both 3DMark's with Virtu enabled; results are below:
3DMark Vantage's performance scarcely twitched with Virtu enabled; there were no signs of visual artifacts or other rendering issues.
Things are much the same in 3DMark 11, although we saw a small drop of about seven percent when we enabled Intel's HD 3000 graphics.We thought we might see problems here since 3DMark 11 uses DirectX 11 while Sandy Bridge's GPU only supports DX 10.1, but nothing of the sort occurred.
Dirt 2 showed a falloff when the GTX 480 was paired with the HD 3000, though not a serious one. The software version we tested was a Release Candidate, not the final version. Issues like this may be entirely resolved by the time Virtu ships out.
The performance difference between the GTX 480 and the GTX 480 with Intel's HD 3000 enabled is well within the margin of error here.
One Significant Annoyance:
The screenshot above (and it's not-shown ATI counterpart) constitute the one thing we dislike about Virtu. While the software is running, there's no way to access NV or ATI control panels—Virtu convinces Windows that it's actually running off the SB GPU. It's unclear if Lucid has a way to eliminate this quirk—the software can be shut off at the task bar, but the display cable must still be shifted back to the discrete GPU.
If you don't mess with the control panel and prefer to define things using in-game controls, there's nothing to worry about here.
|Intel Insider, Conclusion|
We invite all readers who've "enjoyed" DRM over the past ten years to write and tell us how much you loved the experience.
One of Virtu's additional capabilities (we'll leave it up to you as to whether or not this counts as a feature) is that it enables Intel's "Intel Insider" technology to function simultaneously with a discrete GPU. What makes the above particularly amusing is that Intel fell all over itself insisting that Intel Insider was not a DRM technology.
Dubious definitions aside, Intel is planning to roll out its own multimedia content service that SB customers might be interested in.
Intel will launch its own service later this year and hopes that technologies like this will make studios feel better about posting their content online. The only way for anyone to use this service and run a dGPU is to use Virtu—P67 users are out of luck on this front as well.
Right now Virtu looks like a solid product. It can't undo all the issues that prevent enthusiasts from taking full advantage of Intel's new Quick Sync technology but it does at least improve general compatibility. The fact that Virtu is a software-only solution that'll be shipping out with various motherboard OEMs should also help it gain popularity.
If you're weighing the difference between waiting for Z68 and springing for P67, here's how things break down. Z68 will allow for CPU and GPU overclocking (at the moment, P67 is CPU overclocking-only, while GPU-overclocking is reserved for H57.) Z68's bus clock will also be more overclocking-friendly; present Sandy Bridge motherboards can only reach 105-108 percent above stock speeds. Finally, Z68 will allow enthusiasts to use an SSD as a cache drive to speed system performance. Exactly how effective this will prove is up in the air, particularly since the feature may require significant patches from Microsoft in order to reach its full potential.
Despite Z68's benefits, Virtu will still be required to use both the integrated GPU's Quick Sync and a discrete GPU's powerful 3D rendering capabilities. The value of waiting is therefore dependent on how much you care about overclocking as opposed to how much you want an SB-equipped system right now.