When Intel took to the internet to announce Arc last year, it also teased a new super-sampling technology to go along with its new GPUs. In the days following the announcement, the company unveiled XeSS (Xe Super Sampling), Intel’s answer to NVIDIA’s DLSS and AMD’s FidelityFX Super Resolution. We already know the inner-workings of XeSS and how it compares to the other two big names, but today we finally get a release window to look forward to.
Intel just wrapped up its first Arc Graphics event where it unveiled a bunch of discrete mobile GPUs along with teasing a desktop one to come. At the event, we saw XeSS in action again in a working demo of Dolmen, an upcoming sci-fi game, but also announced a launch window for the technology. XeSS will hit the market early this summer, most likely alongside Arc 5 and Arc 7 series GPUs.
As a reminder, XeSS uses machine-learning to upscale your games to display more visual fidelity than native quality. The game is rendered at a lower internal resolution, then upscaled through AI to look better, this results in a higher frame rate with little to no loss in image quality. DLSS does this to great effect where some games actually look worse without DLSS. However, DLSS requires proprietary machine-learning hardware found only inside RTX cards for AI-acceleration, Intel’s XeSS can work just fine with and without it, just like AMD’s FSR.
Intel revealed that more than 20+ games will support XeSS at launch, some of which are:
Intel’s website for XeSS also displays logos of both PUBG Studios and Techland (developers of Dying Light), suggesting that the technology will make its way over to PUBG and Dying Light 2 eventually, as well. Moreover, games that already have DLSS implementation should face little difficulty implementing XeSS as the foundation is already laid by having DLSS support. Here’s what Intel’s Principal Engineer Karthik Vaidyanathan had to say about XeSS implementation as compared to DLSS:
It should be similar and there’s another way to look at it. So for a game that implements TAA already, integrating something like XeSS should only be a small amount of effort because you already have all the pieces that we need with any TAA Implementation. Like you have the motion vectors, you have the jitter. […] TAA has almost become like a de facto, you know, standard for antialiasing. So for any game that already has TAA, it already has the pieces that you would need to integrate XeSS or any super sampling technique with a few modifications, of course, but those are small modifications.
While XeSS is similar to DLSS in many ways, such that it also uses temporal upscaling and works better with hardware-acceleration, it differs from both DLSS and FSR in one key way. Unlike DLSS and FSR’s four available modes, Intel’s XeSS will support 5 different modes: Ultra Performance (3.0x scaling factor), Performance (2.0), Balanced (1.7), Quality (1.5), and Ultra Quality (1.3). That covers a wide variety of applications and puts greater control in the hands of the user.
By the time Intel launches XeSS in summer of this year, AMD‘s FSR 2.0 should also be available conjuring up an interesting competition. AMD switched FSR 2.0 to temporal upscaling, which means now all three upscaling technologies are on the same page, but unlike XeSS, AMD does not require special XMX (Intel proprietary hardware) or DP4a instruction support and it works on rival GPUs as well.
To see Intel XeSS in action, you can click here to watch the most recent demo.
SMART researchers enable early-stage detection of microbial contamination in cell therapy – EurekAlert
Rapid and accurate analytical test method capable of early-stage detection of microbial contaminations enhances the production of high-quality cell therapy products Singapore-MIT Alliance for Research