The NVIDIA Turing GPU Architecture Deep Dive: Prelude to GeForce RTX
by Nate Oh on September 14, 2018 12:30 PM ESTUnpacking 'RTX', 'NGX', and Game Support
One of the more complicated aspects of GeForce RTX and Turing is not only the 'RTX' branding, but how all of Turing's features are collectively called the NVIDIA RTX platform. To recap, here is a quick list of the separate but similarly named groupings:
- NVIDIA RTX Platform - general platform encompassing all Turing features, including advanced shaders
- NVIDIA RTX Raytracing technology - name for ray tracing technology under RTX platform
- GameWorks Raytracing - raytracing denoiser module for GameWorks SDK
- GeForce RTX - the brand connected with games using NVIDIA RTX real time ray tracing
- GeForce RTX - the brand for graphics cards
For NGX, it technically falls under the RTX platform, and includes Deep Learning Super Sampling (DLSS). Using a deep neural network (DNN) specific to the game and trained on super high quality 64x supersampled images, or 'ground truth' images, DLSS uses tensor cores to infer high quality antialiased results. In the standard mode, DLSS renders at a lower input sample count, typically 2x less but may depend on the game, and then infers a result, which at target resolution is similar quality to TAA result. A DLSS 2X mode exists, where the input is rendered at the final target resolution and then combined with a larger DLSS network.
Fortunately, GFE is not required for NGX features to work, and all the necessary NGX files will be available via the standard Game Ready drivers, though it's not clear how often DNNs for particular games would be updated.
In the case of RTX-OPS, it describes a workload for a frame where both RT and Tensor Cores are utilized; currently, the classic scenario would be with a game with real time ray tracing and DLSS. So by definition, it only accurately measures that type of workload. However, this metric currently does not apply to any game, as DXR has not yet released. For the time being, the metric does not describe performance any publicly available game.
In sum, then the upcoming game support aligns with the following table.
Planned NVIDIA Turing Feature Support for Games | |||||
Game | Real Time Raytracing | Deep Learning Supersampling (DLSS) | Turing Advanced Shading | ||
Ark: Survival Evolved | Yes | ||||
Assetto Corsa Competizione | Yes | ||||
Atomic Heart | Yes | Yes | |||
Battlefield V | Yes | ||||
Control | Yes | ||||
Dauntless | Yes | ||||
Darksiders III | Yes | ||||
Deliver Us The Moon: Fortuna | Yes | ||||
Enlisted | Yes | ||||
Fear The Wolves | Yes | ||||
Final Fantasy XV | Yes | ||||
Fractured Lands | Yes | ||||
Hellblade: Senua's Sacrifice | Yes | ||||
Hitman 2 | Yes | ||||
In Death | Yes | ||||
Islands of Nyne | Yes | ||||
Justice | Yes | Yes | |||
JX3 | Yes | Yes | |||
KINETIK | Yes | ||||
MechWarrior 5: Mercenaries | Yes | Yes | |||
Metro Exodus | Yes | ||||
Outpost Zero | Yes | ||||
Overkill's The Walking Dead | Yes | ||||
PlayerUnknown Battlegrounds | Yes | ||||
ProjectDH | Yes | ||||
Remnant: From the Ashes | Yes | ||||
SCUM | Yes | ||||
Serious Sam 4: Planet Badass | Yes | ||||
Shadow of the Tomb Raider | Yes | ||||
Stormdivers | Yes | ||||
The Forge Arena | Yes | ||||
We Happy Few | Yes | ||||
Wolfenstein II | Yes |
111 Comments
View All Comments
Yojimbo - Saturday, September 15, 2018 - link
Of course he meant giga as in billion. Of course it's simple. But where did you see "gigarays/sec" before? You didn't. So as far as you know, it is an NVIDIA-made term.He said it that way because it's easier to talk about it that way, and because NVIDIA is making a marketing term out of it. And he introduced it with a joke.
edzieba - Saturday, September 15, 2018 - link
'Gigs' is an SI prefix. You put it in front of any unit to indicate 10^9 of that unit.eddman - Saturday, September 15, 2018 - link
Why does it even matter?! As I pointed out, I meant to write "rays/sec". I forgot to take out the giga part after copy/paste.I'm pretty sure a lot of ray tracing developers have uttered the words "giga rays" before jensen.
markiz - Monday, September 17, 2018 - link
Well car makers are also using kilowatts for engines. Bastards.Yojimbo - Saturday, September 15, 2018 - link
Where do you get these die area estimates from?edzieba - Saturday, September 15, 2018 - link
If you compare scaled die shots of Turing to Pascal (GP102), isolate an SM, and assume each CUDA core occupies the same die area, then the combination of Tensor and RT cores takes around 24% die area. This is more of an upper bound, as the CUDA cores are likely to be larger due to the concurrent execution capability, there is more space in the SM taken by memory, and more uncore taken by the NVLink interfacesYojimbo - Saturday, September 15, 2018 - link
Where are these die shots? Paste them? I think that people are looking at schematics, not die shots.edzieba - Saturday, September 15, 2018 - link
Nvidia posted them (surprisingly correctly scaled) as part of the Turing unveil presentation.Yojimbo - Saturday, September 15, 2018 - link
Can you link me to them and to an analysis of them?808Hilo - Sunday, September 16, 2018 - link
So far 10% better than 1080ti and 15% over 1080. Really not worth it. Better spend money on MB, new chip, ram and fast SSD if coming from an older PC.