OpenDLSS reimplements Nvidia's DLSS 5 neural rendering network in Vulkan, bit-exact
OpenDLSS: A Vulkan Reimplementation of Nvidia's DLSS 5 Neural Rendering Network
A Vulkan reimplementation of Nvidia's DLSS 5 neural rendering network, matching the original bit-for-bit. The same 71-block Swin/ViT network runs FP8 on tensor cores, with all 75 block boundaries byte-identical. A WebGPU port achieves the same exactness in a browser without tensor cores or FP8. You supply the weights. Performance reaches 2.8 ms at 768x768 on an RTX 4070 SUPER.
The exactness is in the specification, not in the hardware.
- Lerc
I'm rather surprised that it doesn't take the z-buffer as an input. I would have thought that would have provided useful information, it's one of the more useful forms of contolnet.
- TheJCDenton
> bit-exact against the original
What kind of sorcery is this ? Very impressive work !
- tim-projects
Jensen : Nobody needs to code anymore...
Programmer: OpenDLSS...
Jensen : Wait. Not like that! (╯°□°）╯︵┻━┻
- drnick1
> A Vulkan reimplementation of NVIDIA's DLSS 5 Neural Rendering network, bit-exact against the original.
Bit-identical, I swear I heard that somewhere before.
- flohofwoe
Almost 8ms on 1080p resolution seems extremely expensive, does the original also eat into the rendering budget as much?
- binsquare
At what point does this neural rendering take away the human touch on the art styles?
- Geee
What kind of dataset is used to train DLSS 5? Do they need to generate synthetic image pairs first?
- franticgecko3
How useful is this without weights?
Isn't the mote that Nvidia has is they work with studios to generate the training data from the game, then they ship a model per game?
Or is my knowledge outdated here and they're just using a single generalised model?