fbpx

NVIDIA DLSS 2.0: Upgrading real-time ray tracing

Most people know NVIDIA’s latest RTX-series of GPUs for its inclusion of RTX cores that handle real-time ray tracing. Often overlooked, it also includes tensor cores that handle Deep Learning Super Sampling that improves visual fidelity and performance. Recently, NVIDIA has released a newer version of DLSS – DLSS 2.0 that includes several improvements over its first iteration. But before we go deeper, let’s first discuss what DLSS is.

DLSS is an AI rendering technology that boosts frame rates and delivers sharper subjects on PC games, upscaling its output resolution from native rendering. Powered by the NVIDIA RTX TensorCores found on the brand’s RTX GPUs, the technology first rolled out on the RTX 20 series.

NVIDIA DLSS 2.0
The DLSS 2.0 is the latest and upgraded version of the deep learning neural network which accelerates further the performance of the NVIDIA RTX GPUs at all resolutions. This is made possible by applying NVIDIA’s AI and machine learning technology called Neural Graphics Framework to predict future frames, maximize raytracing, and lessen GPU load on graphically demanding games. It now performs twice as fast compared to its predecessor.

Basically, the technology inputs an aliased, jittered pixels and motion vectors together on a special type of AI network called Convolutional Autoencoder to optimize and produce images at the highest resolution. The network also stores the process made on certain frames for mimicking the same details for future renders and improving the stability of the game frame by frame. Players may choose among three options—Quality, Balanced, and Performance—which enables them to adjust the game’s resolution up to 4K.

Among the improvements of the DLSS 2.0 includes great upscaling across all GeForce RTX GPUs and resolutions, application of temporal feedback techniques for producing more detailed renders, customizable image quality modes, and a generalized network that works across each PC game.

NVIDIA’s DLSS 2.0 is featured on specific titles, including the much-awaited PC version of Death Stranding. Directed by Hideo Kojima, the open-world, action game has gained a following due to its rich story and sophisticated graphics.

Under NVIDIA DLSS 2.0, the game has significantly boosted its frame rates across the RTX GPUs. It increases the frame rates by 48% in Quality and 90% in Performance modes. With the DLSS 2.0 turned on for the Quality mode, the RTX 2060 can go up to 71 FPS at 4K resolution. As you can see below, the subjects look sharper, smoother, and more realistic.

Customers may play the Death Stranding with DLSS 2.0 on NVIDIA’s GeForce Now, the brand’s cloud game streaming service, and Game Ready Drivers for the optimum gaming experience.

Until June 29, 2020, customers may purchase select GeForce RTX 20 series GPUs, as well as gaming desktops and laptops bundled with Death Stranding on PC. Those who purchase may redeem the game until August 31, 2020, through a download code.

For more information, you visit this link here.

Subscribe to Blog via Email

Enter your email address to subscribe to this blog and receive notifications of new posts by email.

Join 983 other subscribers
Avatar for Jewel Sta Ana

Jewel is a Multimedia Producer for YugaTech. She's a caffeinator by day and a cinephile by night.

2 Responses

  1. Avatar for Kenneth Kenneth says:

    “barely notice a difference” means it actually works. You don’t seem to have understood what DLSS does so I’ll make an example.
    Let’s say you have game A which supports DLSS. Your pc with a 4k native monitor, can run game A at:
    ~80fps @ 1080p
    ~40fps @ 1440p
    ~25fps @ 4k
    DLSS takes in a lower resolution input (1080p), upscales it to 4k using your gpu’s tensor cores and produces a 4k image output that you can “barely notice any difference” as compared to rendering at native 4k. Now your pc runs game A ~70fps @ 4k.

  2. Avatar for gimmick gimmick says:

    barely notice any difference, plus only visible when not moving. rtx is a waste of money.

Leave a Reply
JOIN OUR TELEGRAM DISCUSSION

Your email address will not be published. Required fields are marked *