Even if the leakers were reliable sources, 5xxx series just too far off (2025) to be taken seriously.
That looks absolutely stunning! Ray construction definitely looks to be the next step and can't wait to see it in action in CP2077 and shortly after in Alan Wake 2.
Nvidia DLSS 3.5 Tested: AI-Powered Graphics Leaves Competitors Behind | Tom's Hardware Now, with Ray Reconstruction, that performance gap is set to become a performance chasm, because not only does Nvidia deliver vastly superior RT performance, but it can also provide clearly superior image fidelity. And Intel's GPUs, which generally have proportionally better RT and AI capabilities than AMD's offerings (though not necessarily in Cyberpunk 2077 2.0), don't get a break either since they're also left out in the cold with the Ray Reconstruction have-nots. It's not a win for everyone, that's certain. What remains to be seen is when the truly killer ray tracing games will start to arrive, how quickly developers will begin leveraging Ray Reconstruction, whether AMD and Intel can offer a competing alternative, and when we'll pass the tipping point where RT performance can't be considered an optional extra. And while all that is going on, Nvidia will likely be researching even more AI-powered "neural rendering" techniques.
NVIDIA DLSS 3.5 Ray Reconstruction Review - Better Than Native - Conclusion | TechPowerUp Today is all about ray reconstruction though. We've done extensive testing with NVIDIA's new tech and I have to say I'm impressed. For years we've seen ray traced reflections that look grainy and smeared out, but we've accepted that as "that's how it has to be, or performance will suffer (even more)." NVIDIA wasn't happy with that explanation, so they thought long and hard about the problem and came up with a thinking-outside-of-the-box solution. Instead of throwing more rays at the problem, Ray Reconstruction takes advantage of additional information, which is available in the game engine, but lost at some point in the pipeline, because subsequent rendering steps work with less information. Take a look at our comparison screenshots—there really is a significant difference in the final output. I do have to admit that especially in motion and during busy gameplay these differences are not very noticeable, but I still think they improve image quality a little bit, which is why it seems appropriate calling this "DLSS 3.5" instead of "DLSS 4.0." It's not just reflections that look sharper, there's substantial improvement to shadows, too, especially when it comes to fine detail. When the rays are traced, they often "miss" fine-structure detail, because there simply isn't enough rays to sample the structure. With Ray Reconstruction, these objects are resolved much better and shadows look sharper and more detailed. We also noticed improvements to ambient occlusion which looks a bit more realistic. Classic denoisers aggregate multiple frames of information, which means some moving objects like the headlights of a car in motion get smeared out—with RR the headlights are crisp and sharp. The same is happening when the source of a reflection flickers quickly—classic denoisers often average out the effect or show it with a short delay. Unfortunately Ray Reconstruction is not supported with DLAA or rendering at native resolution. NVIDIA confirmed that they will be training RR for DLAA in the future. What's even more problematic in my opinion is that right now, Ray Reconstruction only works with Path Tracing enabled. This limits the technology to very few gamers—basically those with RTX 4080 or RTX 4090. Everybody else will be playing with classic ray tracing, or even RT disabled. From a technology perspective, nothing stops Ray Reconstruction from working with classic Ray Tracing effects. At the moment, Ray Reconstruction will be limited to a small audience, but once NVIDIA can solve these two problems it will be a game changer for RT in games. I also think that AMD will be able to implement a similar technology for FSR, and it might even help them gain some extra performance, because fewer rays could be used. An interesting discovery is that Ray Reconstruction actually lowers VRAM usage. For example at 4K we measured 11.8 GB with RR disabled, and only 10.9 GB with RR enabled. One possible theory is that the default denoiser (which gets replaced by RR) has a higher memory usage, possibly because it keeps more history frames in its buffer. What might also help is that RR is integrated with the DLSS Super Resolution pass, which means some buffers can be shared and don't get duplicated.
Spoiler: Interesting DirectX CASO: Significantly more performance for hybrid graphics (IGP plus graphics card) Source: Microsoft September 26, 2023 at 1:15 p.m by Valentin Sattler - A DirectX function called Cross Adapter Scan-Out (CASO) can significantly speed up games when an IGP and a graphics card are used together. This is possible through reduced copying processes, which allows images to be transferred more quickly. Many current processors have an integrated graphics unit, and at the same time a much more powerful graphics card is installed in the system. This combination can be advantageous for users: If you connect the display to the economical, integrated graphics unit, the graphics card can be largely deactivated when idle and thus save energy. However, if a lot of computing power is required, the graphics card simply forwards the calculated images to the other graphics output. CASO reduces copying processes However, this step has so far been suboptimal in the past. In order to display an image, the graphics card first had to copy it into the main memory, where Windows then moved it to another memory area so that it could be displayed. However, this double copying process increases latency and comes at the expense of bandwidth. With functions such as Nvidia's Advanced Optimus or AMD's Smart Access Graphics, there are hardware solutions that can switch the display output internally, but these are not available on all systems. If images from the graphics card are output via the APU's screen output, two copy processes currently have to be carried out. The DirectX function CASO should make the second one superfluous in the future. Source: AMD A function called Cross Adapter Scan-Out (CASO), which has been available for DirectX 11 since 2021 and for DirectX 12 since this year, is intended to solve this problem. According to Microsoft, this makes the second copy process unnecessary and can sometimes significantly increase performance. In one test, for example, AMD speaks of an average of eight percent more fps, and in exceptional cases such as League of Legends, an increase of 50 percent is said to be possible using DirectX CASO. Microsoft goes even further: they are talking about an average of 16 percent more fps and a 27 percent reduction in latency.
Nvidia's Blackwell B100 GPU to Hit the Market with 3nm Tech in 2024: Report Nvidia is set to adopt TSMC's 3nm-class process technology next year, a report by DigiTimes reads. The company is expected to produce its codenamed GB100 compute graphics processor using the technology. However, it is likely that the company will produce all Blackwell GPUs using the same fabrication technology. The report claims that Nvidia's GB100 is projected to be launched in 2024, which will align with Nvidia's typical two-year cadence of launching new GPU architectures... https://www.tomshardware.com/news/n...o-hit-the-market-with-3nm-tech-in-2024-report
I'll wait for reviews anyway. But yeah, might as well get one of those too, if they really are that much of a punch again. Not sure about release price, power consumption (plug!), and chiplet design... I expect those cards to properly work only after updates to windows / drivers / schedulers.
Blackwell B100 GPU: HPC/AI/Compute in 2024. The GeForce GB102 is rumored for 2025. If true, more time for the 4090 to stay the top dog.
Again, not a fan of rumors of products that are still far away from release, but for those inclined to lap it up regardless. https://videocardz.com/newz/nvidia-...d-512-bit-memory-bus-according-to-kopite7kimi
4A Games on next-gen: "We are fully into ray tracing" | Eurogamer.net It's official. 4A Games are fully committed to ray tracing technology for its next big game and the implications are exciting. The firm is in preparing for the next generation with initial work concentrated around RT - strongly suggesting that ray tracing support isn't an afterthought built into next-gen console silicon, but a key feature of the design. In an interview centred on 4A Games' excellent work in bringing Metro Redux to Switch, we asked CTO Oles Shishkovstov for his reaction to publicly revealed aspects of next generation console hardware. "We are fully into ray tracing, dropping old-school codepath/techniques completely," he told us. ... 4A has been at the forefront of working with hardware-accelerated ray tracing. A Metro Exodus demo showcasing the firm's stunning RT-based global illumination technology first debuted running on Nvidia workstation-class Tesla GPUs back at GDC 2018, before the game was demonstrated at Gamescom the same year running on top-tier RTX 2080 Ti hardware. A further pre-launch optimisation push saw the game able to run at 1080p60 with RT enabled at console-equivalent settings on RTX 2060.
I'm so eager to see it. With games like Black Myth Wukong coming next year it will be awesome to play that on next gen hardware. The 4090 will most likely perform well but I do think it may give it a challenge. Mine too fellas. I'm going to sell my current cards and then purchase Blackwell when released. That's the one cool thing about these cards. The resell value helps alot when upgrading.
