After we extensively tested 3rd generation Ryzen processors with the GeForce RTX 2080 Ti, our idea behind this new feature is to add mainstream and budget GPUs to the mix in a benchmark run that reflects more settings and resolutions, gamers will likely use gaming to optimize their PCs. That's why we chose the RTX 2070 Super, the Radeon RX 5700 and the now under $ 200 Radeon RX 580 to get a decent selection of GPUs, which are also our favorites for the Best Graphics Cards 2019 belong.
Instead of just testing the more theoretical 1080p resolution to avoid graphics test bottlenecks, we also added 1440p results. In terms of quality settings, we did not stick to the usual ultra-type settings, but found it expedient to cover ultra, high and medium presets in order to display scaling results.
All in all, we have four GPUs available for testing with two resolutions and three quality presets. This means that 72 benchmark runs are required for each CPU tested for each game tested, with an average of three runs reported for each test. Our three CPUs of choice are two top competitors from Intel and AMD – the Core i9-9900K and the Ryzen 9 3900X – and what we think is the best available CPU, the $ 200 Ryzen 5 3600. With four games selected, we are Agree Near the 900 benchmark run mark, which lasted just over three 16-hour sessions that only benchmarked to get this data – let's share Steve's pain for a second.
The games we chose for this test are Rainbow Six Siege, Assassins Creed Odyssey, Far Cry New Dawn and World War Z. Our test system used the G.Skill FlareX DDR4-3200 CL14 memory with Corsair H115i Pro installed. Automatic overclocking functions such as MCE or PBO have been deactivated, so the results for this hardware are immediately ready for use thanks to the use of high-quality cooling. Let's get to the results!
First, we have the 1080p Ultra Quality World War Z results. With the RTX 2080 Ti we see exactly the same 1% low performance as with the 9900K and 3900X, which is not to be expected due to the results with the RTX 2070 Super and the RX 5700, since the 9900K was 3-4% faster with these slower GPUs. The 9900K was also 4-5% faster when you look at the average frame rate with the 2080 Ti and 2070 Super, and this margin is reduced to just 3% with the RX 5700.
If we then fall back on the RX 580, which still pushes well over 100 fps at 1080p, all three tested CPUs basically deliver the same performance. It's interesting to note that the 3900X with the RX 5700 was 6% faster than the 3600, and that the margin with the RTX 2080 Ti only increased to 8%, since we weren't far from it even with the RX 5700 at 1080p, To be CPU bound.
When switching to 1440p, the RTX 2080 Ti can be easily removed from the RTX 2070 Super and RX 5700. When comparing the data RX 5700 and RTX 2070 Super, however, we see some curiosities. If we look at the 2070 Super, Intel seems to be preferred here because both the 3900X and the 3600 are limited to around 170 fps. In contrast, the RX 5700 allows the 3900X to match the 9900K, while the 3600 drops off here.
With the RTX 2080 Ti, the 3600 reaches its limits in terms of 1% low performance, and interestingly, the 3900X and 9900K are now more evenly matched than with the RTX 2070 Super. As expected, given what we've seen at 1080p, the RX 580 creates a strong GPU bottleneck that neutralizes all margins.
If you lower the preset quality to high, the performance hardly changes compared to the ultra quality, and the edges remain largely the same.
We only see a 5-10 fps drop at 1440p from Ultra to High and again this didn't really affect margins and we still see the same strange scaling between the RTX 2070 Super and the RX 5700.
Even if you use medium graphics settings, the performance is not significantly improved. We only see 10-20 fps gains over Ultra. The margins are all pretty much the same, although the additional headroom on the RX 580 gradually changes performance. For example, the 9900K was 5% faster than the 3600.
The World War Z at 1440p Medium Settings Test shows a similar scaling as the 1440p Ultra results. The most interesting thing is how much the R5 3600 drops when using the RX 5700, which we don't see with the RTX 2070 Super. This is also evident on the RTX 2080 Ti, but these results make a little more sense because the faster GPU enables the faster CPUs to achieve higher frame rates.
Go ahead with some Far Cry New Dawn tests and the first thing we have is the 1080p Ultra results. As you can see, even with the RTX 2070 Super at 1080p, we are very tied to the CPU. The 1% low performance is similar to the RX 5700, but the averages aren't quite as high.
As soon as we use the RX 580, we are almost completely tied to the GPU. As a result, all three CPUs delivered a similar result.
At 1440p, the edges are similar on the RTX 2080 Ti, but they close a bit on the RTX 2070 Super. As expected, there is only a very small difference between the two CPUs in the RX 5700, and we find the same performance again in the RX 580.
With the high-quality presetting, we are once again heavily tied to the CPU, since RTX 2080 Ti and 2070 Super are installed. The situation is similar with the RX 5700. As a result, the 9900K was able to offer 15% more power even with the 5700.
When switching to 1440p, similar margins between 3900X and 3600 are achieved, so that basically the same performance level is achieved with all four GPUs. The 9900K is again 15% faster with the 2080 Ti and 13% faster this time with the 2070 Super, while the margin is reduced to 9% with the RX 5700, and of course the same performance is seen with the RX 580.
With the default quality enabled, we see the RX 5700, 2070 Super and 2080 Ti at 1080p almost the same performance as the 3900X and 3600. In the meantime, the 9900K allows for slight performance gains by moving from the 5700 to the 2070 Super by 6% and from 2070 Super to 2080 Ti by 4%.
The 9900K was up to 24% faster than the Ryzen processors at this low resolution with these mild quality settings, but there was practically no scope to be seen with an RX 580 either.
At 1440p the 9900K was up to 18% faster with the 2080 Ti, 14% faster with the 2070 Super, 10% faster with the RX 5700 and not faster with the RX 580. Keep in mind that this is a worst case scenario for Ryzen. One of the reasons why we wanted to include this game in the test suite.
If we move in the opposite direction, we have one of the best scenarios for Ryzen in Assassin's Creed Odyssey. Here we see the 1% low performance of the Ryzen 5 3600 max with the RX 5700, although we see an increase in the average frame rate as we increase the GPU performance.
With the RX 5700, 2070 Super and 2080 Ti, the 3900X manages to outperform the 9900K for a 1% low and average frame rate.
Increasing the resolution to 1440p largely reduces the margins, although we still see the 3900X advancing with the 2080 Ti for 1% low performance. This time the R5 3600 on the RX 5700 also drops a bit, which we did not see when using the RTX 2070 Super.
If you reduce the quality level to the very high preset, you will achieve a very even performance with the 2070 Super and the RX 580. The RX 5700 seems to prefer the 9900K slightly, and we see a similar thing with the RTX 2080 Ti that is interesting.
When switching to 1440p, the margins remain roughly the same. Here, too, the 9900K with the RTX 2080 Ti offers a very low performance advantage, which is not the case with the 2070 Super or RX 5700.
We skipped the "High" preference because the performance was practically the same as the "Very High" preference, and instead chose the "Medium" preference. Here the R5 3600 with the 2070 Super and the RX 5700 drops a bit at 1080p, which is strange because it didn't have nearly as much to struggle with the RTX 2080 Ti. There is even a very small drop in performance when using the RX 580, although we are only talking about a 4% drop compared to the 3900X.
It gets a little crazy at 1440p. The R5 3600 drops significantly at 1% power, while the 3900X charges before the 9900K when using the RTX 2080 Ti. It is very interesting to see that the 9900K drops from 10 fps faster at 1080p to a few fps slower at 1440p and for the 1% low result it drops quite a bit. We have seen this from time to time in the past, and it seems that in these CPU-demanding titles, the extra load of higher resolution can favor CPUs with more resources. Overall, however, the 3900X and 9900K were evenly matched to one another.
The last game tested is Rainbow Six Siege. First, we have the highest quality results at 1080p, where the 9900K offers a small performance advantage of 4 to 6% when using the RTX 2080 Ti and 2070 Super. This margin has lost almost nothing with the RX 5700, and we are still working over 150 fps on average. Then there are no margins worth mentioning in the RX 580.
If you jump to 1440p, you will see no difference in performance on the RX 580 and the RX 5700. Meanwhile, the 9900K was only 5% faster with the 2070 Super and 4% faster with the 2080 Ti.
If you drop from ultra to very high, you'll see the same margins on the Radeon GPUs, while the 9900K has a slight performance advantage on the 2070 Super and a reasonable performance advantage on the 2080 Ti.
The edges are very similar at 1440p between the ultra and very high presets. The Core i9 9900K offers a performance advantage when you use the RTX 2070 Super or something faster.
By lowering the quality preset to high, Intel's lead is extended with the 2080 Ti at 1080p. Now it is 10% faster than the 3900X for the average frame rate and reaches well over 200 fps. It was also 14% faster when you looked at 1% low power. For 144 Hz players, even the 3600 was good for over 170 fps at all times when using a very fast GPU with most of the load on it.
The edges on the 2070 Super were similar, but once you hit a GPU limit, the edges on the RX 5700 and below are neutralized.
The 1440p Ultra and High results largely show the same thing. The margins with the 2070 Super and the 2080 Ti are slightly expanded in favor of Intel, but there is nothing to write home about, since we can reach over 130 fps with the 2080 Ti regardless of the CPU used.
Put everything together
At the end of the test, we quickly look at the average performance in this 4-game example. With the RX 580, the 9900K was 1% faster than the 3900X, so both delivered within the margin of error and, of course, the same gaming experience with this special GPU, which is not enough for modern high-end CPUs. The 9900K offered a 3% increase in performance with the RX 5700, which still has a GPU limitation scenario.
With the RTX 2070 Super we see a larger gap. The 9900K was 6% faster on average, although we're not sure whether you'll notice the jump from 144 fps to 153 fps. Then the 9900K with the 2080 Ti was 5% faster on average and this time jumped from 161 fps to 169 fps. Needless to say, these two powerful CPUs are ideal for gaming (further reading: Ryzen 9 3900X vs. Core i9-9900K: 36 Game Benchmark).
For those who play with lower quality settings, we saw that the 9900K with the RX 5700 has a 5% performance advantage at medium quality settings. Then it was an average of 7% faster with the 2070 Super and an average of 11% faster with the 2080 Ti.
After a short trial period, we didn't learn anything new, but it covers all aspects of gaming performance, CPU and GPU scaling in a single article.
For the most part, for example, on an RX 5700 or RTX 2060, you won't notice a big difference between the 3900X and the 9900K. With an RTX 2070 Super there is a certain separation depending on the title. For titles like Rainbow Six Siege, where the 9900K was a good deal faster with 1080p settings in the middle, this margin was reduced at 1440p and with Even with the Ryzen 5 3600, which allows average values of well over 144 fps, you have to wondering how important that really is.
If you want to spend less than $ 700 on a graphics card and want to buy a Core i9-9900K or Ryzen 9 3900X just for gaming, this is a lot better with the R5 3600.Only for games, I wouldn't recommend the 9900K or 3900X, unless you spend $ 1,000 on an RTX 2080 Ti.
The Ryzen 5 3600 is really the right processor at the moment. We even recommend skipping the 3700X for gaming only, as high-performance 8-core / 16-thread processors won't be able to be fully used in games anymore, and the $ 130 you save now if needed for an upgrade can actually be used many years later.
- AMD Radeon RX 5700 XT on Amazon, Google Express
- AMD Radeon RX 5700 on Amazon, Google Express
- GeForce RTX 2080 Ti on Amazon, Google Express
- GeForce RTX 2070 Super on Amazon, Google Express
- GeForce RTX 2060 Super on Amazon, Google Express
- AMD Ryzen 9 3900X on Amazon, Google Express
- AMD Ryzen 5 3600 on Amazon, Google Express
- Intel Core i9-9900K at Amazon, Google Express
- Intel Core i5-9600K at Amazon, Google Express
- Intel Core i5-9400F on Amazon, Google Express
Masthead credit: 3D rendered CPU concept by Blue Andy