Finding the Best CPU for the Nvidia Geforce RTX 2080Ti & 2080 Super
How much does the CPU performance in games from Full HD to 4K resolution differ in comparison, if current processors are accompanied by the currently fastest graphics card in the form of the Geforce RTX 2080 Ti?
Update: In the meantime, we could add the announced measurements with the Core i7 2600K from 2011 to the benchmarks. With this CPU, the fps doesn’t differ significantly between Full HD and WQHD, so the 2600K limits the performance of the graphics card in Full HD in almost all cases.
On average, the Sandy Bridge processor still achieves just over 100 fps in Full HD. The Core i9 9900K’s lead in this resolution is 80 percent, both with the GTX 1080 Ti and the RTX 2080 Ti. In WQHD the 9900K’s lead drops to 52 percent, in 4K only 25 percent remains (in each case in combination with the RTX 2080 Ti).
Clearly larger distances are to be registered against it with the 99th-Percentile-fps, which we supplemented likewise in the article (completely at the end). They give the lowest fps value in the benchmark, but without considering the one percent of the lowest values. This serves to filter out short-term slips that could distort the picture.
Here, the 9900K with the RTX 2080 Ti still has a lead of 44 percent over the Core i7 2600K in 4K, even in 4K, and almost 90 percent in Full HD. The Core i7 8700K with six cores has a lead of 34 percent (4K), respectively 76 percent (Full HD). The performance doesn’t only turn out significantly higher with the current eight-core CPU, but also significantly more even than with the more than seven-year-old quad-core (which isn’t surprising).
One way to improve the 2600K’s performance is overclocking. Measurements for this can be found in our detailed test of the 2600K CPU. With a GTX 1080 Ti, we were able to reduce the gap to the Core i7 8700K in Full HD from 53 percent to 34 percent by increasing the clock rate by 1,000 MHz at the 99th percentile fps, and from 50 to 30 percent at the average fps.
When we create a new CPU test system, we usually use the fastest graphics card on the market at the time. This is to ensure that the GPU limits the performance of the different processors as late as possible.
While the Geforce GTX 1080 Ti held the performance throne for a long time, the Geforce RTX 2080 Ti is now a clearly faster graphics card. We’ve taken this as an opportunity to take another look at the Core i9 9900K, the Core i7 9700K and the Ryzen 7 3700X. We’re also planning additional measurements with the Core i7 2600K later on to show how well a CPU that is over seven years old performs in this comparison.
Fresh benchmarks paired with the Geforce RTX 2080 Ti should clarify the question of whether, and if so how much, the distance between the processors increases due to the significantly faster GPU. At the same time, we also increase the number of used resolutions: In addition to Full HD (1920×1080), we also measure in WQHD (2560×1440) and in 4K (3840×2160).
While we deliberately leave these higher resolutions out of the regular CPU test system, as the graphics card becomes more and more the limiting factor with increasing resolution, we have added them in this case to show exemplarily how strong the aforementioned limitation actually works in practice.
Since a more current Geforce driver is used for the measurements than in the case of the regular test system, the determined values differ slightly from the previous results. Wolfenstein 2 is the only game that makes a comparatively large leap in performance due to the newer driver, but overall the differences are only about one percent on average.
Test Results: Best CPU for Geforce RTX 2080Ti & 2080 Super
Ranking First: Intel Core i9 9900K
- Sets a new bar for single-GPU performance
- Quiet, cool-running design
- Easy to attain at least modest overclocks
- Supports ray tracing and DLSS for future games.
Best performing CPU for the RTX 2080Ti & 2080 Super
There’s no question, the Core i9 9900K from Intel proves to be an extremely fast CPU in the test, as expected, whether in games or applications. However, there are the odd bland aftertaste – and that’s completely independent of the questionable benchmarks, which Intel itself published in advance.
These certainly include the price and the delivery situation. You can’t really buy the Core i9 9900K at the moment, whereas the cheapest listing is currently 650 Dollar. This is about twice as much as the predecessor Core i7 8700K was available for at best times (which is currently also only poorly available and considerably more expensive than usual).
Furthermore, the Core i9 9900K runs at its limit in terms of clock rates, temperatures and power consumption, according to my test experiences, already from factory condition – and that despite the long awaited change from thermal paste to soldered metal between CPU die and heatspreader.
All in all, this gives the impression that Intel wanted to send an eight-core with very high clock rates into the mainstream race by hook or crook in order to steal the show from the Ryzen 7 3700X in the best possible way.
Verdict: Best performing CPU for the RTX 2080Ti & 2080 Super
Since the predecessors of the 9900K are also comparatively expensive at the moment, AMD can still sit back and let Intel do its job quite easily. In terms of performance, Team Blue is often slightly ahead, especially in games. In practice, this is relevant enough to justify the current price difference to processors with comparable core and thread numbers from Team Red. All in all, if you want the best perfromance with an 2080Ti or 2080 Super, you need the Intel i9 9900K CPU. Our perfromance winner.
Ranking Second: AMD Ryzen 7 3700X
- Best price-performance ratio
- Great memory and cache performance
- Improved overclocking margins
- Lots of motherboard choice
- Limited overclocking headroom
Best price-performance ratio CPU for the RTX 2080Ti & 2080 Super
With Zen 2 alias “Matisse” AMD says goodbye to the old Zeppelin die structure on the chip and splits the tasks into several parts: Up to three components sit on the silicon of the Ryzen 3000 series. Two of them are so-called chiplets. These are the nuclei of the Ryzen – a maximum of eight per chipplet, divided into clusters of four. The cache near the CPU is also located in the chipplets. Both chipplets communicate with the IO-Die via the “Infinity Fabric” data bus (“IO” stands for “in/out”). This in turn takes care of data transfer to the rest of the PC, memory management and also transfers information between the chipplets.
The chipplets are also the epitome of AMD’s current pride in the number 7 – which is why the processors also appear on 7.7. The cores are manufactured with a structure width of 7 nanometers. The Ryzen 2000 series was still produced with 12 nanometers. The miniaturization allows a CPU manufacturer to either shrink a die to make it more efficient or to put more processing units on the same surface.
For the user, the new structure does not change anything at first. In fact, Matisse brings hardly anything new in terms of pure functionality – except that the AMD CPUs are the first to introduce PCI-Express 4.0 into consumer platforms. However, the performance advantages of the wider data bus (512 instead of 256 bits) are currently manageable for normal users – for example, graphics cards are not expected to exhaust the gigantic bandwidth in the near future. But there’s nothing wrong with air to the top. And PCIe 4.0 SSDs can be faster than their older counterparts.
By the way, the CPU still fits into the AM4 socket and can be overclocked with most motherboard chipsets. So if you are still satisfied with your first-generation Ryzen board and the manufacturer offers updates, you can still stick with your old board.
Get it all out
With a price point of around 350 Dollar, you should think that the direct competitor would be the Intel Core i7-9700K. After all, both CPUs have eight cores, whereby the 3700X supports simultaneous multithreading and thus 16 threads simultaneously, whereas the Intel processor doesn’t. However, the benchmarks show that not only the current flagship – the Ryzen 9 3900X – but also its little sister CPU is boxing against the Intel Core i9-9900K with eight cores and 16 threads. The performance is often at least on par, and often the 130 Dollar cheaper CPU even wins.
In comparison, highlights can be seen in the everyday simulation PCMark 10 (R7: 4,150 points, i9: 3,800 points), raytracing (R7: 4,350 points, i9: 4,250 points) and synthetic gaming benchmarks with the support of an Nvidia GTX 1080. AMD scores 20,200 points in 3DMark Fire Strike, Intel only 19,900 points. In 3DMark Time Spy the difference is even a bit bigger: 8,050 points against 7,700 points.
In many tests, however, the two CPUs are almost equal. Intel takes the speed victory in large spreadsheets and compresses data faster. The i9 is also clearly in front in single thread mode in the Cinebench render benchmark.
But as already mentioned: at this price difference, it is a fantastic performance that the AMD chip can keep up with Intel. All benchmark results of our tests can be found in the following table.
|AMD Ryzen 7 3700X||Intel Core i9-9900K|
|PCMark 8||4,194 points||4,152 points|
|PCMark 10||4,165 points||3,783 points|
|Excel||0.449 seconds||0.41 seconds|
|Cinebench R15||2,171 points||2,033 points|
|Cinebench R20||4,948 points||4,912 points|
|Cinebench R20 (ST)||499 points||511 points|
|Winrar||22.935 KB / s||25.476 KB / s|
|Handbrake||166.7 FPS||157.51 FPS|
|x264||118 FPS||120.3 FPS|
|x265||10.9 FPS||10.2 FPS|
|POV-Ray||4,335 points||4,273 points|
|TrueCrypt||697 MB / s||697 MB / s|
|Fire strike||20,238 points||19,899 points|
|Time Spy||8,067 points||7,681 points|
Ryzen 7 with good efficiency
The Ryzen 7 is also not compromised in terms of power consumption. In the benchmark suite PCMark 10, our test system needs 232 watts or 331 watts, depending on the scenario. The extended test is especially interesting here, as the overall performance is thus 17 watts below the Core i9 and 19 watts below the Ryzen 9.
You have to keep in mind, however, that we have to use different mainboards per manufacturer, of course, which in turn influence the power consumption. All in all, it can at least be said that AMD hasn’t saved on efficiency.
The secret is in the IPC
There is a big difference between AMD and Intel: the clock frequency. While Intel has cracked the 5 GHz, the Ryzen Boost just manages to get to 4.4 GHz. So the strong performance can only come from a monstrously improved IPC (Instructions per Cycle). AMD lists a few changes here that, taken together, should contribute to the additional 15 percent IPC that the manufacturer claims compared to the previous generation.
Most obvious is the increased L3 cache. 32 MByte of CPU-like memory are now available. The improved AVX2 support is also exciting – the CPU now processes corresponding data twice as fast. Furthermore the chip improves the jump prediction of instructions, gets a larger micro-op cache and a more associative L1 cache.
The last two improvements are a bit more vivid: first, the so-called thread grouping. In Zen 2, processor threads, i.e. tasks of the executing programs, prefer to end up in the same chiplet and there in the same computing cluster rather than at different ends of the processor. This might be a better solution especially for the spatially separated chipplets.
Memory: AMD has given the Infinity Fabric, i.e. the CPU data bus, more freedom in terms of clocking. This should remove an old bottleneck – nevertheless, according to AMD there is a “sweet spot” with DDR4-3733. If you want to save a little money without significant performance losses, you should go for DDR4-3600 (CL16). Unfortunately, we haven’t been able to test how different data rates affect the performance so far.
Verdict: Best price-perfromance ratio CPU for the RTX 2080Ti & 2080 Super
There is one thing you shouldn’t forget when it comes to CPU performance: AMD does without an integrated graphics unit in the higher class desktop processors. If Intel were to omit this, there would be more space available for CPU tasks. However, an integrated graphics unit can sometimes also bring significant advantages in benchmarks. All in all the Ryzen 7 3700X CPU is a great choice for the RTX 2080Ti & 2080 Super and because the price is more than fair, the Ryzen 7 3700X wins our Best price-performance ratio award in Best CPUs for the RTX 2080Ti & 2080 Super.
Ranking Third: Intel Core i7 9700K
- 8-cores / 16-threads
- Soldered IHS
- Great gaming performance
- No bundled cooler
Great performing CPU for the RTX 2080Ti & 2080 Super
With the ninth generation of the Intel Core series, Intel introduces the refresh of the Coffee Lake architecture. This in turn was a refresh from Kaby Lake, which in turn was a refresh from Skylake – so we have been waiting in vain for real innovations since 2015. Intel is thus still stuck with the 14-nanometer (nm) process technology, while AMD is moving to 7 nm next year. The i9-9700K does bring some improvements, but the advantages over its predecessor are very limited.
|Intel Core i7-9700K||Intel Core i9-9900K||AMD Ryzen 7 2700X||Intel Core i7-8700K|
|Base clock frequency||3.6 GHz||3.6 GHz||3.7 GHz||3.7 GHz|
|Boost clock frequency||4.9 GHz||5.0 GHz||4.35 GHz||4.7 GHz|
|L2 cache||8 x 256 KB||8 x 256 KB||8 x 512 KB||6 x 256 KB|
|L3 cache||8 MB||16 MB||16 MB||12 MB|
|PCMark 10 Extended||3,689 points||3,703 points||3,864 points||3,587 points|
|Microsoft Office 2016: Excel||2.2 seconds||2.1 seconds||2.6 seconds||2.6 seconds|
|Cinebench R15||1,522 points||2,017 points||1,823 points||1,377 points|
|Handbrake 0.9.5||111.7 fps||134.5 fps||128.3 fps||111.0 fps|
|x265 benchmark||8,692 fps||9.272 fps||7.72 fps||8.02 fps|
|PovRay 3.7 RC3||3,584.73 pixels / s||4,022.03 pixels / s||3,696.91 pixels / s||2,976.71 pixels / s|
|TrueCrypt AES Twofish Serpent||492 MB / s||695 MB / s||624 MB / s||483 MB / s|
|3DMark Fire Strike (GTX 1080)||19,899 points||19,864 points||18,732 points||19,338 points|
|3DMark Time Spy (GTX 1080)||7,498 points||7,681 points||7.906 points||7,451 points|
Only slightly faster than the i7-8700K
The Intel Core i7-9700K does well in all benchmark tests and hardly shows any weaknesses. Both in the everyday benchmark, PCMark 10, and in elaborate Excel simulations and video coding, the processor is almost on par with the Intel Core i9-9900K and the AMD Ryzen 7 3700X. Understandably, the performance isn’t quite as good in strongly scaling applications like Cinebench – in comparison to the two top processors, it lacks hyperthreading. This affects for example the Cinebench rendering benchmark or encryptions in TrueCrypt.
The predecessor, the Intel Core i7-8700K (6 cores, 12 threads) is outperformed by the Intel Core i7-9700K (8 cores, 8 threads) in all tests, but often only by a hair’s breadth. There’s nothing to complain about in the gaming area, though. In tests of the 3DMark suites, the CPU in combination with the Nvidia GTX 1080 graphics card achieves almost identical rates to the Intel Core i9-9900K.
Z390 chipset: Everything in it
Also the updates of the new mainboard chipsets are not without. While Intel with the Z370 (and most other 300 series chipsets) and the LGA1151v2 socket cheekily forced a board update without significant improvements, the new Z390 actually brings a breath of fresh air. USB 3.1 Gen 2 with up to 10 GBit/s is now natively supported, and WLAN-ac and Bluetooth 5 are also on board. The latter features have to be explicitly built in by the motherboard manufacturer, however, and could drive up the price of these boards a bit.
The extra should cost buyers less than previous components with integrated wireless features. As a result, customers will have much more choice from various USB and WiFi combinations. All ninth generation CPUs are compatible with all 300 series chipsets.
Verdict: Great performing CPU for the RTX 2080Ti & 2080 Super
The Intel Core i7-9700K is a great CPU for work and gaming PCs, especailly when using a Geforce RTX 2080Ti & 2080 Super. In the test, the processor achieves good benchmark results overall. Thanks to high clock rates, many cores and overclock potential, you won’t go wrong with this processor. However, the i7-9700K is also quite expensive, only slightly better than its predecessor and is clearly surpassed by the i9-9900K.
CPUS for RTX 2080Ti & 2080 Super – Game benchmarks in detail
Assassin’s Creed: Origins (High graphic details settings)
If the fps don’t decrease (or only decrease to a very small degree) despite increasing resolution, in the case of our test system one can assume that the processor limits the performance. An extreme example of this is the graphically less demanding Civilization 6: No matter which combination of CPU and GPU we use, there are no significant differences in performance between Full HD and 4K. But the situation is different when comparing the processors.
The i7 8700K in Civilization 6 is about 20 percent faster than the Ryzen 7 3700X and the 9900K is about ten percent faster than the 8700K. This applies to both the RTX 2080 Ti and the GTX 1080 Ti. On the other hand, this doesn’t really play a role in practice with refresh rates in the range of 150 fps and more.
In the other benchmarks, the fps usually fall clearly with increasing resolution. In some cases, there is hardly any difference between Full HD and WQHD, at least with the Ryzen 7 3700X, which suggests the mentioned limitation by the processor.
This applies primarily to Assassin’s Creed: Origins, Kingdom Come and Project Cars 2, whereby the limitation in the benchmarks only stands out due to the extremely high performance of the Geforce RTX 2080 Ti. But since we’re again dealing with (almost) three-digit fps values, this limitation is easy to get over in practice.
Nevertheless, the Intel processors profit much more from the RTX 2080 Ti’s additional performance in Full HD. Whilst the Ryzen 7 3700X only shows noteworthy performance increases in Total War: Warhammer 2 and in Wolfenstein 2, the 9900K and the 8700K, with the exception of Civilization 6, always increase, and in some cases significantly.
The performance rating shows a total performance increase of 15 percent in the case of the Core i9 9900K in Full HD with the RTX 2080 Ti in comparison to the measurements with the GTX 1080 Ti. With the Core i7 8700K it is eleven percent, whereas the Ryzen 7 3700X only achieves a plus of five percent.
As expected, the Intel CPUs can therefore increase their distance to the Ryzen 7 3700X a good bit in Full HD thanks to the RTX 2080 Ti. However, on the one hand it has to be considered that we “only” measure in high instead of maximum details, which means that the GPU limit only takes effect much later, since maximum details in some games put a disproportionately high load on even the RTX 2080 Ti (in view of the mostly only marginal visual improvement). On the other hand, the differences between the CPUs in higher resolutions are significantly reduced.
Whilst the Core i9 9900K in Full HD with the RTX 2080 Ti is 32 percent ahead of the Ryzen 7 3700X, it is 17 percent in WQHD and only ten percent in 4K. And with maximized graphics details, the difference would very likely turn out even smaller or even non-existent, as the graphics card alone limits the fps.
Conclusion of the editors
Our new CPU measurements with the Geforce RTX 2080 Ti confirm that Intel is (still) a step ahead in games – at least when the graphics card doesn’t become the limiting factor. However, the higher the resolution and graphics details (and the slower the graphics card used), the closer the processors move together.
This is more relevant than ever, especially considering the current price and delivery situation of Intel processors. Because if the price differences were smaller (contrary to the current situation), it would be more tempting to go for the fastest model, even if the performance differences rarely play a really noticeable role in practice.
Significantly higher prices, on the other hand, also represent a greater hurdle for the purchase, which AMD is currently playing into its cards, since Ryzen CPUs are considerably cheaper and readily available. And when it comes to performance in games, it should not be forgotten that developers are getting to know the Zen architecture better and better over time, which could reduce the performance differences in the future.
Considering that the Core i9 9900K is already pushing Intel’s current architecture to its limits (see our review of the 9900K) and given the ongoing difficulties with production in 10 nanometers, I’m very excited to see what Intel’s next steps will be.
For gamers, the good news is that both AMD and Intel currently offer plenty of cores and more than enough CPU power for gaming (and many other tasks) – even if prices and availability, especially for Intel models, leave a lot to be desired.
Overview: The Nvidia RTX 2080Ti
In the run-up to Gamescom, Nvidia presented the first three Geforce graphics cards from the Turing family: The RTX 2080 Ti as a new top model based on the 754 mm² TU102 graphics chip and below that the RTX 2080 with TU104 GPU as well as the RTX 2070 (probably also TU104, possibly TU106). Nvidia has already released the first two graphics cards for pre-ordering, which is why all major board partners have presented their first custom designs. We’ll give a first insight into the partly unusual solutions at this point. In the coming weeks the article will be supplemented with new information.
Custom designs of the RTX 2080 Ti: More cooling capacity, but almost no final specifications
It is noticeable that in most cases no clock rates and no power limits have yet been set. Merely EVGA, Gainward and Palit give rates which are oriented on Nvidia’s reference specifications, respectively the slightly increased rates of the Founders Edition – as is generally known, Nvidia provides the latter with a “factory OC” of a nominal 90 MHz boost clock and an additional 10 watts in the power limit. PNY names the highest energy budget of 285 watts so far, whereas it could be a mistake at least in the model with a Direct Heat Exhaust-cooler including radial fan.
Service for you: Maximum power limits
With the step-by-step arrival of the RTX custom designs the test after the test starts for us – we compare the board partners’ own designs with Nvidia’s Founders Edition. The first models have already passed test runs, so that we can submit the public specifications around clock rates with our own knowledge. The most important information for passionate overclockers concerns the maximum power limit. This determines how well a Geforce RTX 2080 Ti can hold its boost during overclocking. The maximum power limit is a value above the factory set 100 percent. For example, the Founders Edition works with 260 watts ex factory and can be set to up to 320 watts in tuning tools (123 percent). The output value is important when considering this, because 125 percent on a 250 watt card is less than 110 percent on a 300 watt bolide.
The rule “more is better” also applies to Turing, but makes good cooling all the more important. Rumor has it that Nvidia has set 350 watts as upper limit, but no partner has set this value so far. The previous peak values are 330 watts (MSI Gaming X Trio), respectively 338 watts (EVGA XC Ultra). Upcoming top models could exhaust the maximum value.
Even if the final power limits haven’t been set in most cases yet, the custom designs make it clear that the manufacturers have set themselves higher values. After the 2.5-slot coolers became more and more popular with the Geforce GTX 1080 Ti, the first manufacturers dare to take the step to three slots with the RTX 2080 Ti. MSI’s Geforce RTX 2080 Ti Gaming X Trio is the most impressive Turing graphics card to date: Three axial fans are accommodated over a length of 327 mm (possibly 2 × 100 mm + 1 × 80 mm), with a thickness of 55.6 mm it occupies almost three slots, and weighs a total of 1,870 grams. In addition, the power supply is realized via two 8-pin and one 6-pin connector. We wouldn’t be surprised if the RTX 2080 Ti Gaming X Trio with the OC BIOS ended up at 300 watts. As a smaller version, MSI will also offer the Duke 11G OC in a 2.5-slot design directly in Germany this time.
Inno3D is launching its black model for the first time with a 240 mm dual radiator instead of a 120 mm model. In addition to the TU102 GPU, the AiO water cooling keeps the voltage converters and the GDDR6 memory at temperature, so that the manufacturer does without the additional fan on the graphics card. EVGA’s Geforce RTX 2080 Ti XC Ultra Gaming could potentially belong to the quieter custom designs: With three slots, the cooling solution turns out massive, but the standard power limit should still be 250 watts.
Asus’ ROG Geforce RTX 2080 Ti Strix is still missing
Worth mentioning: The two sister companies Gainward and Palit are increasingly going their separate ways. Whereas Gainward now relies on three axial fans, Palit stays with two. Meanwhile, Asus has only presented its turbo and dual models of the Geforce RTX 2080 Ti. A ROG Strix is likely to take on the competitors’ flagships in the future. Meanwhile, Gigabyte still holds back its Aorus top model and only shows the smaller Windforce OC and Gaming OC. At the other end of the spectrum, the manufacturers are again offering solutions with DHE coolers, which should approach Nvidia’s “Ab” UVP of a good 1,000 Euros. Partly, the boards of the Founders Edition are used there, recognizable by the Nvidia logo on the PCI Express connector.