Digital Foundry Delivers A Technical Breakdown Of Switch 2’s DLSS
By We’ve been aware that Nvidia’s specifically designed Switch 2 chip would introduce some DLSS (Deep Learning Super Sampling) enchantment to the upcoming console for some time now, although specifics on how the tech actually operates behind the scenes have been quite scarce. Thankfully, the technology specialists at Digital Foundry are available to illuminate the entire topic.
In a recent video (above), DF’s Alex Battaglia delves into the details of the Switch 2’s DLSS and clarifies how it compares to other variants on PC. There hasn’t been any official announcement from Nintendo or Nvidia regarding this issue, but this is virtually as thorough an analysis as one can find.
There’s a plethora of technical jargon to comprehend in the aforementioned video, and we strongly recommend viewing it fully for all the particulars. However, to summarize: there are two distinct types of DLSS present on Switch 2.
The first is what we observe implemented in titles like Cyberpunk 2077: Ultimate Edition and Street Fighter 6 on Switch 2. According to Digital Foundry, this DLSS type shares numerous similar traits with the Convolutional Neural Network (CNN) variant, which was prevalent on PC before the launch of DLSS 4. This offers more stable image quality at 1080p resolutions and functional anti-aliasing on both movement and camera transitions.
Nonetheless, this isn’t equivalent to all DLSS types available on Switch 2. Analyzing titles such as Star Wars Outlaws, Hogwarts Legacy, The Touryst, and Fast Fusion on the console, Digital Foundry deduces that there exists another, “tiny” DLSS model for games striving to achieve resolutions higher than 1080p.
This secondary model features a clearer image resolution but significantly poorer anti-aliasing on movement and camera transitions — the DF analysis even mentions that it appears ineffective during motion. The major advantage is that this variant could function at nearly half the energy requirement of the more powerful version, reducing the strain on the Switch 2’s GPU for upscaling.
Of course, this is solely based on DF’s evaluation for the moment, without any official statements from Nvidia or Nintendo (as improbable as it would be to receive anything from the latter). That said, Digital Foundry has communicated with a developer, who remains unnamed, that the theory is accurate, and there are indeed two DLSS types in action for the system.
Battaglia wraps up his analysis by expressing that he understands why developers might prefer to utilize the less demanding model in handheld mode, though its visual compromises lead him to question its application in docked mode.
What are your thoughts on the Switch 2’s DLSS functionalities? Is it a crucial aspect of your gaming experience? Share with us in the comments.