Monday, March 18th 2019

NVIDIA to Enable DXR Ray Tracing on GTX (10- and 16-series) GPUs in April Drivers Update

NVIDIA had their customary GTC keynote ending mere minutes ago, and it was one of the longer keynotes clocking in at nearly three hours in length. There were some fascinating demos and features shown off, especially in the realm of robotics and machine learning, as well as new hardware as it pertains to AI and cars with the all-new Jetson Nano. It would be fair to say, however, that the vast majority of the keynote was targeting developers and researchers, as usually is the case at GTC. However, something came up in between which caught us by surprise, and no doubt is a pleasant update to most of us here on TechPowerUp.

Following AMD's claims on software-based real-time ray tracing in games, and Crytek's Neon Noir real-time ray tracing demo for both AMD and NVIDIA GPUs, it makes sense in hindsight that NVIDIA would allow rudimentary DXR ray tracing support to older hardware that do not support RT cores. In particular, an upcoming drivers update next month will allow DXR support for 10-series Pascal-microarchitecture graphics cards (GTX 1060 6 GB and higher), as well as the newly announced GTX 16-series Turing-microarchitecture GPUs (GTX 1660, GTX 1660 Ti). The announcement comes with a caveat letting people know to not expect RTX support (think lower number of ray traces, and possibly no secondary/tertiary effects), and this DXR mode will only be supported in Unity and Unreal game engines for now. More to come, with details past the break.
NVIDIA claims that DXR mode will not run the same on Pascal GPUs relative to the new Turing GTX cards, primarily given how the Pascal microarchitecture can only run ray tracing calculations in FP32 mode, which is slower than what the Turing GTX cards can do via a combination of FP32 and INT32 calculations. Both will still be slower and less capable of hardware with RT cores, which is to be expected. An example of this performance deficit is seen below for a single frame from Metro Exodus, showing the three different feature sets in action.
Dragon Hound, an MMO title shown off before by NVIDIA at CES, will be among the first game titles, if not the very first, to enable DXR support for the aforementioned GTX cards. This will be followed by other game titles that already have NVIDIA RTX support, including Battlefield V and Shadow of the Tomb Raider, as well as synthetic benchmarks including Port Royal. General incorporation into Unity and Unreal engines will help a lot coming forward, and NVIDIA then mentioned that they are working with more partners across the board to get DXR support going. Time will tell how well these implementations go, and the performance deficit coming with them, and we will be sure to examine DXR in the usual level of detail TechPowerUp is trusted for. But for now, we can all agree that this is a welcome move in greatly increasing the number of compatible devices capable for some form of real-time ray tracing, which in turn will enable game developers to focus on implementation as well now that there is a larger market.
Add your own comment

113 Comments on NVIDIA to Enable DXR Ray Tracing on GTX (10- and 16-series) GPUs in April Drivers Update

#101
FordGT90Concept
"I go fast!1!11!1!"
Makes sense, double the flops for slight loss in accuracy. Pretty sure Radeon Rays has 16-bit precision options.

Edit: github.com/GPUOpen-LibrariesAndSDKs/RadeonRays_SDK/blob/master/RadeonRays/src/intersector/intersector_lds.cpp
if (spec.has_fp16)
m_gpudata->qbvh_prog.executable = m_device->CompileExecutable("../RadeonRays/src/kernels/CL/intersect_bvh2_lds_fp16.cl", headers, numheaders, buildopts.c_str());
Yup, RadeonRays SDK has FP16 checks.
Posted on Reply
#103
FordGT90Concept
"I go fast!1!11!1!"
The FP16 code paths wouldn't be taken on Polaris at all. They aren't something the GPU understands.
Posted on Reply
#104
HwGeek
What about FP8/INT8? can they used somehow in RTRT?
Posted on Reply
#105
FordGT90Concept
"I go fast!1!11!1!"
No, DXR would likely be implemented using FP32 if AMD were to add backwards compatible support for it. I assume that it would just be noisier (fewer rays/bounces).
Posted on Reply
#106
Vlada011
Than I didn't wrong because paid 580 euro for GTX1080Ti Poseidon before 5 months???
Peformance of RTX2080 FE, 11GB, option for water cooling (not full cover waterblock) but at least for that money is included cooler comparable with AIO kits worth over 100$.
And option to install full custom waterblock for ASUS Strix when price drop to around 70 euro.

We can say different things for AMD, but they help to everyone with their research and trying to offer for lower price.
Intel and NVIDIA alone would massacred us with their politic and prices without AMD.

First impression after GeForce users switch to high end Radeon graphic card will be... I feel like picture quality is little better, like sharper cleaner more photographic image.
And in future if NVIDIA continue to ask 1000$ for high end GPU and AMD offer 10% weaker for 750$, I'm ready to consider combination Intel-Radeon.
Posted on Reply
#107
jabbadap
FordGT90ConceptNo, DXR would likely be implemented using FP32 if AMD were to add backwards compatible support for it. I assume that it would just be noisier (fewer rays/bounces).
Well there might be some use of int cores with in it. Not necessary as low precision as int8 is but that exodus frame pic shows quite hefty use of int32 math during RT(Possibly just filtering/denosiong but anyhow). I think the possibilty to do fp32 to and int32 math concurrently will give one way to do it hard way. Volta and Turings can do it that way and Pascal obviously can't. Which bodes the question: how is GCN, can it run fp32 and int32 math concurrently?

Posted on Reply
#108
Maximuspop
MAXLDThat's the point of this move. Making the "poor" look not so poor in comparison.

"- Boys, let's release this RTX we've been working on for a few years, we'll have it exclusively in the new cards, so we'll sell them for premium.
- Boss, it's not that great yet, our new hardware still isn't that capable of fast and proper implementation.
- Just do it, it will be the first hardware Ray Tracing bling-bling ever, it's a big deal. We'll get it working in a couple AAA games and people will jump into it.
(... few months later...)
- Boss, people aren't joining the RTX bandwagon... and they aren't really swapping Pascal for Turing.
- Well then execute plan B: unlock Ray Tracing for old Pascal.
- But boss, those have no Ray Tracing focused hardware, it will run tons even worse.
- Exactly, we'll make them feel that Pascal is ancient crap, and then they'll want to finally swap them for RTXs. At the same time we'll spread the name even more.
(... weeks later...)
- Boss, still not a big interest in RTX 2000 cards. What now?
- Fine, release the RTX 3000 series with proper improved RTX performance, we'll make RTX 2000 look like ancient crap in comparison, and RTX 3000 look like the second coming of baby Jesus."


Just like they said when they presented RTX, bringing a card to the market that can do Ray Tracing this "fast" (compared to before) is quite an achievement. Problem is: it's still not fast enough.

"- People, we made the impossible: a card that can finally do the legendary tech that is "Ray Tracing"! Behold!
- Cool!
(...) Ok, nevermind, it runs slow. And I can live without it for now, can barely see the difference anyway.
- No, you don't understand. This is dope engineering. If it wasn't for this new hardware, it would be a slideshow with your current card.
- Y, but it's still slow. Not appealing.
- Look, we'll show you. Let's test with your current card.
- Dude, please..."
:clap::rockout::laugh::laugh::laugh::laugh::laugh:
Posted on Reply
#109
SoNic67
jabbadapVolta and Turings can do it that way and Pascal obviously can't. Which bodes the question: how is GCN, can it run fp32 and int32 math concurrently?
I doubt that Turing can do it 100% either, even if they say "independent integer execution units", I think it's like Hyperthreading in CPU's. I might be wrong though...
Posted on Reply
#110
jabbadap
SoNic67I doubt that Turing can do it 100% either, even if they say "independent integer execution units", I think it's like Hyperthreading in CPU's. I might be wrong though...
Well it can do fp32 at full throttle and concurrently run some int operations with it's separate int cores. It's actually all about keeping gpu busy doing floating point math.

On Turings whitepaperthey say even on current games they could get some performance benefit only from that. And is the one of the main reasons why Turing gets more performance out of TFlops compared to Pascal(The second being new shared cache).
Posted on Reply
#111
SoNic67
Yes, I saw that. This is the part reminded me of hyper-threading concept on CPU's:
First, the Turing SM adds a new independent integer datapath that can execute instructions concurrently with the floating-point math datapath.... This translates to 2x more bandwidth and more than 2x more capacity available for L1 cache for common workloads.
Is not like they have added extra integer units. To me it looks like they can now issue those INT instructions in the same time with FP32 ones.

Later this gets a little more... confusing.
Two SMs are included per TPC, and each SM has a total of 64 FP32 Cores and 64 INT32 Cores. In comparison, the Pascal GP10x GPUs have one SM per TPC and 128 FP32 Cores per SM. The Turing SM supports concurrent execution of FP32 and INT32 operations (more details below), independent thread scheduling similar to the Volta GV100 GPU.
So in the first sentence they say that those are separate cores. But the numbers (64FP+64INT) add to the same number like on Pascal (128FP). And the last sentence talks again specifically of "thread scheduling"...
I don't think there would be a physical difference between an INT32 and a FP32 core. It's only the schedulers that make that difference.
Overall, the changes in SM enable Turing to achieve 50% improvement in delivered performance per CUDA core.
This is in line with the hyper-threading gains on CPU's.
Posted on Reply
#112
jabbadap
SoNic67Yes, I saw that. This is the part reminded me of hyper-threading concept on CPU's:

Is not like they have added extra integer units. To me it looks like they can now issue those INT instructions in the same time with FP32 ones.

Later this gets a little more... confusing.

So in the first sentence they say that those are separate cores. But the numbers (64FP+64INT) add to the same number like on Pascal (128FP). And the last sentence talks again specifically of "thread scheduling"...
As it says Turing TPC has _two_ of those (64int+64fp) SMs, Pascal TPC have _one_ 128FP SM. On the one more confusing note: according to Nvidia Turings without tensor cores have separate fp16 cores.
Posted on Reply
#113
SoNic67
Hmm...
GP102-450-A1 has 3840 cores and 30 Streaming Multiprocessors. 3840/30=128
GP104-410-A1 has 2560 / 20 = 128
TU102-400-A1 has 4608 cores and 72 Streaming Multiprocessors. 4608/72=64
TU104-400-A1 has 2944 / 46 = 64

To me looks like definitelly they broke those SM in two and they can issue either INT32 or FP32 commands to them, on independent paths. Which is consistent with approx 50% gains.
Posted on Reply
Add your own comment
Nov 22nd, 2026 16:44 CST change timezone

New Forum Posts

Popular Reviews

TPU on YouTube

Controversial News Posts