قالب وردپرس درنا توس
Home https://server7.kproxy.com/servlet/redirect.srv/sruj/smyrwpoii/p2/ Technology https://server7.kproxy.com/servlet/redirect.srv/sruj/smyrwpoii/p2/ AMD Radeon VII Anti-Aliasing & 5K, 8K Benchmarks vs RTX 2080 | PlayersNexus

AMD Radeon VII Anti-Aliasing & 5K, 8K Benchmarks vs RTX 2080 | PlayersNexus



This content is a functionally academic exercise, which means that we sometimes go out of "real-world" uses rather than exploring how the card behaves and rinses at the basic level. This means looking at irresistible scenarios, such as 8K resolution, to see if zooming improves AMD through NVIDIA; in theory, based on what AMD told us about productivity, we should confirm it. Anti-aliases take multiple samples per pixel, often 2x, 4x, 8x or 16x, and the effect is very similar to the increasing resolution.

If the bandwidth of AMD memory is really good for gaming, what we think would be a fall or change in performance between Radeon VII and RTX 2080 as resolutions and anti-aliasing, with a worsening or increasing gap, more favored by RTX when the distinction is reduced ability. We have already shown how Radeon VII can sometimes solve a bigger resolution when we published our review, and we talked about how the gap is like in 201

8. F1, will increase by 1080p. Now is the time to see if it cancels or closes high resolutions; however, keep in mind that the obvious drawback is that we enter an area where content is not reproduced in a frame, so it's called academic activity.

Superposition AA Testing

Starting from the shadows set to "Low", "High" and the resolution of 4K, 5K, and 8K resolution are found, we saw that NVIDIA took 27.4 percent. Leading the benchmark, moving to 27.9 percent 8K. Here, in fact, there is no change from 4K to 8K, at least as far as scales are concerned, although it is obvious that these resolutions are getting smaller as they grow.

We went on a one-off test with Extreme Shadows to determine that AMD really got it in this scenario. AMD had an average of 1566 points, and NVIDIA had 1504 points. This test is re-tested four times and averaged to determine that this delta is outside the standard deviation

After the greater differences with Extreme shaders, tests are performed from 1080p to 4K. At 1080p, the NVIDIA RTX 2080 performed about 25.8% better than the Radeon VII card at 100% on the original line. At 1440p, the advantage of the RTX 2080 quickly dropped to 16.1%. The 4K results are too close and create an inconvenient chart, but the RTX 2080 actually surpassed Radeon VII and Radeon VII a couple of percentage points. Now, with reference to the "real" number of frames, the benchmark worked at around 11-12 FPS AVG. It is not playable, clearly, and this advantage is really incomprehensible, because no one actually uses these settings, but it is interesting to see how two cards can flip positions in different test scenarios.

GTA V AA Testing [19659004] GTA V has many anti-aliasing options, so we will use both MSAAs from 2x to 8x and then enable 8x reflections in MSAA. We will only start with the AVG FPS chart with different settings, which will best show how the difference between each test changes

4K and very high and very high settings everything, 2x MSAA presents us 58.8FPS AVG for 2080 and 51.4FPS AVG Radeon VII, but the more important indicator is that in 2080 has 14.4 percent In the case of 4x MSAA, this leader does not change much and decreases to 14.1% by mistake. At the beginning of the 8x MSAA we see the movement. AVG FPS is 27.5 Radeon VII and 31FPS RTX 2080, and the RTX 2080 has improved to 13% compared to the original 14.4%. Then also adding 8x MSAA reflection contrasts, we see that the difference is 10.8% advantage 2080 FE. Seeing the line chart in total, we can see that AMD is recovering some of the losses to the extreme end of the scale just because we are out of reach for other reasons

. We passed the standard deviation of about 0.3 FPS AVG. These results are very accurate

Firestrike AA Testing

Firestrike remains one of the best tools for this synthetic workload. That's why we will separately review GT1 and GT2. The GT1 heavily recharges the GPU with polygons and tessellation, shades and illuminates, and uses a computing shadow after processing and particle physics. The GT2 is very difficult to calculate the shadow workload and greatly increases the number of single-point-treated points, but reduces the workload of tessellation by more than 50%. Therefore, GT2 should be better than Radeon VII than GT1,

Starting with GT2, we see that Firestrike Ultra settings have NVIDIA below AMD performance, 86.7 percent. This is unique for this benchmark. 2x MSAA, this gap is reduced to 90.7% of baseline. The 4x MSAA brings us 94.6% of the base line, while the 8x MSAA ultimately allows the RTX 2080 to exceed AMD and maintain 104.7% of its initial operation.

It is important that AMD is usually the most competitive GT2, where the workload and memory count is so high. Because of the fact that NVIDIA is starting to move forward with MSAA 8 times, we are working with the 3DMark team and others to better understand the specific behavior. In our estimation, this may be related to memory compression using NVIDIA, or possibly edge detection using anti-aliasing with this particular software installation, or other implementation-level advantage.

GT1, where tessellation and polygons are heavier, we see that two Cardestrike Ultra cards are starting to work just like NVIDIA has achieved a significant 16% 2xMSAA, 21% – 4x MSAA, and 27% – 8x MSAA.

8K with 8x MSAA reaches really inappropriate levels of use, but finally see Radeon VII along the front of the RTX 2080. That's what it takes to completely empty the RTX 2080 memory, and then it started choking because it brought and dropped out of memory [19659003] Far Cry 5 HD Texture + AA Testing

Next to Far Cry 5. For this purpose we tried with SMAA, TAA and HD textures according to different scenarios

For the first 4K test and using HD texture with SMAA, we saw NVIDIA at 62.8FPS AVG and AMD 60.6FPS AVG. about 3.5%. 1% lower values ​​were similarly arranged as 0.1%.

To test the scale with different settings, we also tested with reduced VRAM, but the settings were otherwise identical. Now we can go to this chart. The results, when disabling HD textures, were functionally identical to HD textures. Not surprisingly, we experience a margin of error, because textures do not work effectively unless VRAM becomes a constraint. Quantitative experience in both devices has not changed from the previous chart. Qualitatively, each unit has the same output. Again, NVIDIA is about 3.3% faster here.

Switching to TAA, each device reduces performance, but NVIDIA still maintains a 3.3% advantage over AVG FPS. HD Textures and Anti-Aliasing Changes Far Cry 5 is not enough to reduce or change the delta performance of a card from card to card.

Conclusion: Not All We Wanted

Our hypothesis was not entirely correct. We seem to have killed him in a few cases, but Firestrike dropped us off the trail. We've reached the 3DMark team and the NVIDIA engineer so we can better understand the specific scenario. One case where we saw a huge increase in Radeon VII when we exploited the RTX 2080 VRAM, which was made using Firestrike Ultra 8K resolution and 8x MSAA. This is, of course, completely unsuccessful, so we reiterate that this is just a research exercise, not one that is particularly practical. 8K / 8xMSAA saw that Radeon VII operated at about 4.7-4.8FPS AVG until 2080 ~ 1-1.7FPS AVG. This is a "big" percentage improvement and is because 2080 We've exhausted the framebuffer, but it's obvious that it can't run on both devices. Nevertheless, this is interesting data and could help to better understand the other features of the card in future tests

Editorial: Steve Burke
Video: Andrew Coleman


Source link