Shulou(Shulou.com)11/24 Report--

A month ago, CTOnews.com reviewed Nvidia GeForce RTX 4090. Its huge performance improvement and the AI power of DLSS 3 have undoubtedly changed the rules of the game in the GPU world. In addition to RTX 4090, a RTX 4080 16G was also released at the press conference, which is also based on NVIDIA Ada architecture, making all aspects of data more balanced and more suitable for high-end gamers.

CTOnews.com also received the RTX 4080 16G public graphics card in advance. In this test, we will still fill up all the configurations to see if this high-end graphics card can give full play to the power of the NVIDIA Ada architecture. The specific configuration is as follows:

Appearance design: still cold, more compact compared to RTX 4090's powerful appearance, GeForce RTX 4080 16G gives people a sense of restraint of "suit thugs", it does not do the exaggerated shoulder-jumping design like RTX 4090, the whole is a standard 3-slot graphics card, the compatibility for the chassis will be better.

The TGP power consumption of GeForce RTX 4080 16G is 320W. In order to carry higher power consumption, the power supply interface is changed to 16 Pin ATX 3.0. three 8-Pin to 16-Pin adapters are included in the package. One less than the 4 8Pin-to-16Pin that comes with RTX 4090, so the demand for power is not so great. For silent frequency, a full power supply of more than 750W is enough.

The GeForce RTX 4080 16G overall design inherits almost the same design as the RTX 30 series public edition GPU, with dense metal cooling armor and two huge fans in the silver-and-black armor. Vertical air ducts are formed one after the other.

The GEFORCE RTX logo at the top also glows when lit.

Like the RTX 4090, because the weight of the video card is relatively heavy, a screw hole is provided on the side of the video card, and the video card bracket can be installed by opening the cover plate.

The GeForce RTX 4080 16G interface is still standard with 3 DP 1.4 x HDMI 2.1, which can support displays with a resolution of up to 8K.

Technical analysis: core energy efficiency skyrocketing, helping AI leap into our hands of GeForce RTX 4080 16G is based on the current positioning of the second highest AD103 core, there are seven groups of GPC, of which four are complete. There are 9728 CUDA cores, 304 Tensor cores and 112 ROP units inside, which is similar to that of the previous generation RTX 3090.

The core frequency of the GeForce RTX 4080 16G is 2210-2510MHz, which is also a huge improvement over the previous generation. Video memory is achieved in one step to the 256-bit bit width of 16GB GDDR6X. The large-scale improvement of video memory makes 4K games no longer worry about exploding memory, while also taking into account the current popular AI productivity work. The factory preset maximum design power consumption is 320W, which is consistent with the previous generation of RTX 3080 10G version. Considering its huge performance improvement, the energy-efficiency ratio of this generation of RTX 4080 should be quite explosive. We will conduct in-depth testing in the following part of the grill.

The architecture of the GeForce RTX 4080 16G is, of course, the latest NVIDIA Ada architecture, which is based on the TSMC 4N NVIDIA custom process, resulting in a leap in performance and power consumption of up to 2x. The throughput of streaming multiprocessors is twice as high as that of the previous generation. The effective ray tracing computing power of the third generation RT Cores is 2.8 times higher than that of the previous generation. The fourth generation of Tensor Cores adds a new FP8 engine with up to 1.32 petaflops of Tensor processing performance, which is five times higher than that of the previous generation. SER brings up to three times the performance improvement for RayTracing, and up to 25% improvement for overall game performance.

As for the concrete function, the new architecture brings a lot of new functions: due to the introduction of Ada optical flow accelerator, DLSS 3 can predict the motion changes between frames in the scene, realize AI frame insertion, and improve the frame rate while maintaining the image quality. At the same time, it supports the application of the eighth generation NVENC double AV1 encoder, which can not only shorten the video export time, but also have better picture quality.

Theoretical performance: the running score almost doubled and the improvement was amazing. We mentioned earlier that the energy-efficiency ratio of RTX 4080 is excellent, so what is the actual test result? We immediately carried out the theoretical test, the first is a single-baked doughnut, after 15 minutes, the core temperature is stable at 64.6 ℃, the power consumption is stable at about 310W, in line with 320W TGP. The most important thing is that the fan is far from fully loaded, only running about 45%, so it is quite quiet to use. In practical use, a full power supply of 750W is absolutely sufficient if equipped with a processor like i7 / i9.

In the actual game, it is far below the 320W power consumption of TGP. We use the Nvidia FrameView tool to summarize the actual power consumption diagrams of several representative games, which are basically located around 270W to 300W in the game. It seems that the energy efficiency ratio of the core of RTX 4080 is quite excellent.

Next, there is a 3DMark pressure test, which can detect whether the performance of the graphics card has declined under continuous running minutes. Generally, more than 97% can be regarded as a qualified graphics card. The score of the measured GeForce RTX 4080 16G is 99.6%, and the performance release is extremely stable.

In the 3DMark TimeSpy DX12 test, the score of GeForce RTX 4080 16G graphics card reached 26552, compared with 17306 for GeForce RTX 3080 10G, 18981 for RTX 3090, and 21862 for RTX 3090Ti. In other words, the GeForce RTX 4080 16G surpasses all the graphics cards of the previous generation of RTX 30 series and increases by 50% compared with the previous generation of RTX 3080 of the same level.

In the 3DMark Fire Strike DX11 test, the score of GeForce RTX 4080 16G graphics card reached 53379, compared with 40932 for GeForce RTX 3080 10G, 46045 for GeForce RTX 3090 and 52630 for GeForce RTX 3090 Ti. This score means that it can also run all games smoothly at 4K resolution, and it can challenge the high number of frames with the blessing of DLSS 3.

In the 3DMark Portal Royal light pursuit test, the GeForce RTX 4080 16G scored 17250 points, which was improved more than the performance of the grating. It is about 50% higher than GeForce RTX 3080.

The performance of the RTX 4080 16G in the theoretical running score is absolutely shocking. We have made a comparison table of GPU physical scores, and the RTX 4080 16G has a significant improvement compared to the previous flagship card in all events.

DLSS 3 details: AI power game, frame number 4 times soar although the theoretical performance has surpassed all RTX 30 series graphics cards, but the really big improvement is DLSS 3 technology. DLSS 3 is a performance multiplier driven by AI, powered by the new fourth generation Tensor Core and GeForce RTX 40 series GPU's new optical flow accelerator, is the latest version of NVIDIA's acclaimed deep learning supersampling technology, and features groundbreaking optical framing technology that will usher in a new era of NVIDIA RTX neural network rendering games and applications.

The optical flow accelerator based on NVIDIA Ada Lovelace architecture can analyze two consecutive game images and calculate the motion vector data of objects and elements from frame to frame without the need for traditional game engine modeling. The frame generation model acquires both motion vector and optical flow data in the game engine, which greatly reduces the visual anomalies of AI when rendering elements such as particles, reflections, shadows and lights.

To put it simply, in the past, DLSS 2 improved the image quality and the number of frames by rendering a low-resolution image, and then AI zoomed in to high resolution. While the new DLSS 3 is compatible with DLSS 2, it also adds a frame generation function, which can insert a new frame calculated by AI between two real frames to double the number of frames. Combined with the super-resolution function of DLSS 2, AI can reconstruct up to 7/8 display pixels, and the game performance can be up to four times better than without DLSS!

Next two, let's move on to the testing of DLSS 3. The latest version of 3DMark has been adapted to DLSS 3's BenchMark. We can directly see the difference between the test results of GeForce RTX 4080 16G and the control variables of DLSS 2. In DLSS 2 mode, the average frame number is 97.9 frames, which is nearly three times higher than that without DLSS 2.

But if we turn on the latest DLSS 3 technology, the frame number of GeForce RTX 4080 16G graphics cards goes straight to an average of 125.84 frames, a significant increase compared with the number of frames without DLSS.

Only a month later, DLSS 3 has been adapted to a number of new games, and when GeForce RTX 4080 was released on November 16, there were 10 DLSS 3 games on the market:

1. The Legend of the Plague: Requiem (A Plague Tale: Requiem)

2. "memory of Light: infinite" (Bright Memory: Infinite)

3. "destroy all mankind 2: re-detect" (Destroy All Humans! 2-Reprobed)

4. "Shadow Torch City" (F.I.S.T.: Forged in Shadow Torch)

5. F1 ®22

6. "against the cold" (Justice)

7. "reincarnation of Life and death" (Loopmancer)

8. Marvel Spider-Man: remaking (Marvel's Spider-Man Remastered)

9. Microsoft simulated Flight (Microsoft Flight Simulator)

10. "superhuman" (SUPER PEOPLE)

As long as it adapts to DLSS 3, it can be downwards compatible with DLSS 2, and the development difficulty is very low, so it is certain that more and more developers will join the DLSS camp. We selected a number of games already adapted to DLSS 3, such as Cyberpunk 2077, Plague Legend: Requiem, Super Human, Shadow Torch City, memory of Light: infinity and reincarnation of Life and death, to test the GeForce RTX 4080 16G.

We came up to do the most stressful "Cyberpunk 2077". We ran its own BenchMark under the 4K highest picture quality super light, and got the following data. It can be seen that in the case of DLSS 3 performance file, up to 50% increase in the number of frames can be achieved, which is quite significant.

So will DLSS 3 have a serious impact on picture quality? When we compare the pictures with NVIDIA ICAT tool, we can see that there is indeed a difference in picture quality between the left and right sides, but it is not obvious macroscopically, it can be said that DLSS 3 technology really allows us to "pick up" a lot of frames for nothing.

↑ left DLSS balanced, right DLSS 3 off

In "Super Human", a new book by PUBG developer Blue Hole, options are also available for DLSS 3. As can be seen from the data below, DLSS 3 is similar to cyber punk 2077 in "superhuman", with a frame rate increase of nearly 70% after turning on DLSS 3. Directly from being able to play smoothly at 4K to being able to play smoothly at 4K high frames.

Next is a domestic cyber punk game "reincarnation of Life and death". Although it is a 2D horizontal version of the game, but with the light chasing this revolutionary function. It can simulate real panoramic light and reflection in the virtual world. In the picture, the wet ground after the rain can best show the charm of light chasing Boppunk. If you take a closer look at the puddle, the difference is obvious. If the light pursuit is not turned on, there is only a vague projection, and when opened, it will produce a lifelike reflection.

↑ left RT upscale, right RT closed

In terms of frame number performance, DLSS 3 can also bring a frame number increase of up to 2.5 times.

The author found that in the first batch of game libraries of DLSS 3, the number of domestic games is much larger. In addition to the "reincarnation of life and death" we tested above, there are "memory of Light: infinite", "Shadow Torch City" and "Phantom Tower". In memory of Light: infinity, DLSS 3 brings about a 2.5-fold improvement at most.

Another domestic steampunk game, Shadow Torch City, only provides DLSS 3 automatic / off two gears, but the increase in the number of frames is still quite significant, with a maximum performance improvement of about 2.5 times.

Finally, we tested an exploding 3A masterpiece, the Legend of the Plague: the Requiem. The game provided DLSS 3 support when the author evaluated RTX 4090, but because the game is still in beta, we can't successfully complete the test. Now officially launched, the RTX 40 series graphics card can finally be played freely.

After the actual DLSS 3 is turned on, the number of frames can be increased up to 3 times at most, which is very useful for this kind of 3A stand-alone masterpiece with complex materials.

So is DLSS 3 technology perfect? Of course not. The essence of DLSS 3 is that AI generates new frames, and one of the side effects of generating frames must be increased latency. After all, our mouse is the feedback on the frame we see, while AI adds additional frames, which has a great impact on competitive games. So when you turn on the frame generation feature of DLSS 3, NVIDIA Reflex will force it on to minimize the rendering delay caused by the generated frame. According to the actual measurement, the delay after turning on DLSS 3 is lower than that of turning on nothing.

Game measurement: enjoy 4K e-sports, higher frame number and lower delay in addition to supporting DLSS 3 masterpiece, CTOnews.com has also tested several mainstream games for everyone. GeForce RTX 4080 16G this session will be tested at the highest resolution of 4K. In the newly launched "Watch Pioneer" and "return", it provides NVIDIA Reflex low latency option, which enables CPU to start to submit rendering work to GPU before the completion of the previous frame, thus greatly reducing or even eliminating the rendering queue, which makes the response time faster and aiming accuracy higher. After the enhanced Reflex is turned on, the delay is reduced by half.

In Tomb Raider: shadow, it is measured that running BenchMark at the highest resolution of 4K can run up to 137frames, which is basically enough for 4K144Hz monitors to play.

The game is also supported by DLSS 2, which can increase the number of frames by up to nearly 90% after turning on DLSS 2.

Finally, we tested a light chase 3A masterpiece "Control". The highest picture quality and highest light in 4K can hardly be played smoothly without DLSS, while the number of frames of GeForce RTX 4080 16G has almost doubled after opening the initial DLSS, which is an excellent experience.

Generally speaking, GeForce RTX 4080 16G can meet all the requirements of Guangzhuo 3A at 4K resolution. Of course, 16G video memory is still dangerous when playing 8K games. It can be said that the RTX 4080 is a high-end game card that is very suitable for matching 4K144Hz monitors. Let 4090 challenge the 8K games that represent the future.

Creative production: AI era computing power is productivity this generation of GeForce RTX 4080 memory has improved a lot, so it has also been given a certain productivity attribute. There will be some achievements in dealing with creative production and deep learning. At the same time, GeForce RTX 4080 16G can also install NVIDIA Studio drivers, and more than 100 creative applications can call the AI computing power of NVIDIA CUDA core to greatly improve work efficiency.

In modeling and rendering software, GeForce RTX 4080 16G can use CUDA to provide hardware acceleration for work. At present, almost all modeling software will optimize the GPU of NVIDIA, so it can achieve the coexistence of efficiency, stability and compatibility. For example, in the commonly used rendering tool V-Ray, you can use RTX accelerated ray tracing to achieve high-performance final frame rendering. With the noise reduction feature of AI, GPU can further speed up interactive rendering.

In V-Ray Benchmark, RTX 4080 scored 3060, making it possible to cope with most of the modeling rendering needs.

Thanks to the improved performance of light pursuit, the score of GeForce RTX 4080 16G in V-Ray GPU RTX has also increased to 4155, which is enough to cope with high-pressure rendering of real-time light pursuit.

Finally, let's take a look at another rendering software, Octane, which can achieve fast interactive rendering through RTX accelerated ray tracing and OptiX noise reduction with AI. According to the final test results of the GeForce RTX 4080 16G, its rendering capacity is equivalent to about 10 GTX 980s working at the same time, and the improvement in computing power from GTX 980s to RTX 4080 is like a lifetime away.

In the later part of the video, the GeForce RTX 4080 16G is also equipped with the eighth generation NVIDIA NVENC encoder and supports AV1 dual encoders. AV1 is the next generation video coding technology, which can provide faster video coding and higher quality streaming media transmission performance while occupying the same space.

NVIDIA works closely with a number of industry partners, and today's mainstream multimedia applications can support AV1. At present, the domestic professional version of editing and projection has supported the acceleration of NVIDIA's AV1 video encoder. OBS LVB software also supports AV1-encoded recording of NVENC. It is hoped that AV1-encoded LVB can be adapted as soon as possible, thus greatly saving LVB bandwidth.

Commonly used video color editing software DaVinci and clip Professional Edition also provide support for AV1 coding. As a free and open source coding, AV1 will certainly be the mainstream coding in the future.

Why is AV1 favored by manufacturers? Because it can greatly reduce the video volume on the basis of maintaining the picture quality. Let's do a test here. The professional edition is used to suppress the 30000Kbps video of two 4K60 frames at the same time, one using H.264 coding and the other using AV1 coding. As a result, the AV1 version has only 342mb, which has shrunk by 36%.

So will there be any damage to the picture quality? We select a picture in the video and use the NVIDIA ICAT tool to look at the details. As a result, there is almost no difference between the image quality on the left and that on the right. Small and high-definition, no wonder AV1 will become the next generation of industry norms.

CTOnews.com tested PugetBench For Davinci for everyone. GPU acceleration effects based on NVIDIA CUDA technology can achieve faster real-time video editing and frame rendering, and the output speed using NVIDIA coding can be increased by up to 5 times. The final score of GeForce RTX 4080 16G can reach 2570, 8K video clips will not stutter, and there is no need for real-time playback.

In addition, NVIDIA also provides NVIDIA Broadcast as an artifact for RTX GPU. Through artificial intelligence technology, it realizes some practical effects, such as audio noise reduction, green screen matting, portrait focus and so on. For anchors who want to buy a high-end graphics card, with the efficient coding of NVIDIA Broadcast+AV1, full-link LVB can be completed with only one machine.

Conclusion: from the perspective of positioning and pricing, the successor of the flagship of the RTX 30 series, the goal of the RTX 4080 16G is quite clear, which is to replace all the high-end cards of the RTX 30 series to become the flagship of the next generation of 4K games. In our test, it does achieve the highest picture quality of 4K and can eat all the 3A masterpieces. If DLSS 3 is supported, it is possible to achieve 4K high-frame games. We can see that only one month from the 4090 launch, 10 DLSS 3 games have been adapted, and there will certainly be more games to support in the future. And in our actual use, the power consumption of RTX 4080 is always more than 200W, which can be said to be quite outstanding. The author is already looking forward to the game book with RTX 40 series Laptop GPU.

In terms of price, the suggested retail price of GeForce RTX 4080 16G Founder Edition is 9499 yuan. It can also be seen from the increase in prices that the positioning of RTX 4080 is actually similar to that of the previous generation of RTX 3090, both for high-end players who want to experience 4K144Hz, or enterprise users who rely on GPU for a living. The author believes that we can wait for AMD 7900 series and RTX 4080 series graphics cards to appear on the market, compare the price and performance of the two, and then consider whether to buy. But if you are an ordinary player, in fact, the performance of RTX 4080 is also overflowed, I think you might as well look forward to RTX 4060, RTX 4070, or directly start with the RTX 30 series models whose prices are now stable.

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.