I need the brilliance and dark power!
- Status
I need the brilliance and dark power!
Reading about Displaced micro-mesh (DMM) on techpowerup article really catches my interest.
It's no continuous LOD solution, but discrete proxy meshes with displacement might work well enough in practice, and i could easily generate the data.
I wonder how they intend to expose this. Extensions to DirectX and Vulkan?
Or NVAPI, so AAA only?
It's no continuous LOD solution, but discrete proxy meshes with displacement might work well enough in practice, and i could easily generate the data.
I wonder how they intend to expose this. Extensions to DirectX and Vulkan?
Or NVAPI, so AAA only?
GeForce RTX 4090 Performance Figures and New Features Detailed by NVIDIA
NVIDIA shared the first concrete performance figures as well as extra details of the new features available in the GeForce RTX 4090.
wccftech.com
Last edited:
Nvidia is usually very good about offering Vulkan extensions, but with DirectX I really don't know how that would work.
Make it a 'potential future feature', like Intels Traversal Shaders? : /
Anyway, I'd be fine with Vulkan Extensions.
NVIDIA Micromesh
A graphics primitive for micro triangles that can encode opacity or displacement.
developer.nvidia.com
Across platforms, APIs and IHVs. I guess it probably works with D3D, otherwise what APIs are they referring to?
davis.anthony
Veteran
How is DLSS 3 frame rates going to be declared in comparison videos?
Currently you might have:
3840x2160 output (2560x1440 CB to 4k) at 60fps
Now imagine DLSS 3:
3840x2160 output (2560x1440 DLSS Quality) at 112fps (56fps actual fps DLSS 3 to 112fps)
Those poor video overlays.
Currently you might have:
3840x2160 output (2560x1440 CB to 4k) at 60fps
Now imagine DLSS 3:
3840x2160 output (2560x1440 DLSS Quality) at 112fps (56fps actual fps DLSS 3 to 112fps)
Those poor video overlays.
With Direct3D extensions, their "official backdoor" to explicitly accessing more functionality is NVAPI which also includes features like being able to use reBAR memory ...
So the SDKs for displaced micro-mesh and that other opacity one probably require NVAPI and that's how it's all handled.
Until we know the real world efficacy of frame generation these benchmarks have no validity to me.GeForce RTX 4090 Performance Figures and New Features Detailed by NVIDIA
NVIDIA shared the first concrete performance figures as well as extra details of the new features available in the GeForce RTX 4090.
GeForce RTX 4090 Performance Figures and New Features Detailed by NVIDIA
NVIDIA shared the first concrete performance figures as well as extra details of the new features available in the GeForce RTX 4090.
'Only' a 2x increase in normal raster. Its a huge leap considering 'moores law is dead'.
Flappy Pannus
Veteran
When I read the piece and saw those reported temps at 50-57C with full clocks, my first though was was 'well ok guess that's over ambient', but with the chart showing the 3090ti at it's expected 70-75C by comparison, apparently not. That is an astoundingly low temp range for a modern GPU running flat-out, even before counting in DLSS 3.
Of course it's omg xboxone huge, but for a card this expensive and power hungry, I'll give Nvidia kudos for at least seemingly over-engineering the cooling solution. It seems they could have gotten away with maybe even a 2.5 slot cooler but if those temps are accurate, it at doesn't look like it will be difficult to keep these cards very quiet under full load. So points for that.
Of course it's omg xboxone huge, but for a card this expensive and power hungry, I'll give Nvidia kudos for at least seemingly over-engineering the cooling solution. It seems they could have gotten away with maybe even a 2.5 slot cooler but if those temps are accurate, it at doesn't look like it will be difficult to keep these cards very quiet under full load. So points for that.
60% isn't 2x.
@Flappy Pannus That latency is a little tricky because I'm guessing "native" has reflex disabled and DLSS3 has reflex enabled. Maybe I'm wrong, but I think that's likely. If I had to guess with reflex on the native is probably a bit lower than the dlss.
Subtlesnake
Regular
So if I'm reading that right, it looks like they roughly doubled native RT performance in Cyberpunk with the 4090. That doesn't seem super impressive given some of the expectations generated in the announcement. Though obviously achieving similar performance with the Overdrive mode is impressive.GeForce RTX 4090 Performance Figures and New Features Detailed by NVIDIA
NVIDIA shared the first concrete performance figures as well as extra details of the new features available in the GeForce RTX 4090.
Flappy Pannus
Veteran
When people say "normal raster", they're usually referring to actual, non-reconstructed frames. That chart is 2X over native using DLSS 3 performance, the 2X is just for games that don't emphasize ray tracing. As we have no idea the ultimate quality of DLSS 3 yet, difficult to say it's a "2x increase" in performance. I mean yeah, Flight Sim could definitely show doubled frames - but if it has additional artifacts over DLSS 2 in the process, then is it really 'doubled' if ultimately the end resolution isn't up to snuff?
Actual 'normal raster' increases that have been published so far point towards 50-70% increases, and that's for the 4090 only.
@Flappy Pannus I think he means native vs native against the 3090ti. It's actually 168%, not double, but looking at the numbers I can see how it looks roughly double.
I think most of (including FE) the RTX 4090's cooler is designed for >600W overclocking headroom. So within the default TGP range it shows great cooling efficiency.
It's great for FPS charts for marketing though (the same for FSR etc.)
- Status
Similar threads
