Today we're testing the new Intel enthusiast and gaming flagship CPU, the Core i9-10900K. Not to be confused with the Core i9-10900X, a senseless 10-core Cascade Lake-X part for the LGA2066 socket. The 10900K is a much more interesting product with the brand new Z490 chipset with LGA1200 socket support and a cheaper boot price.
The Core i9-10900K is a 10-core processor with 20 threads, a base frequency of 3.7 GHz and a single-core turbo of 5.3 GHz with the new Thermal Velocity Boost from Intel. Compared to its predecessor, the 9900K, the L3 cache size has increased from 16 MB to 20 MB and the TDP from 95 W to 125 W. Like the 9900K, the 10900K has the same MSRP of $ 488, though there are some issues we'll be addressing towards the end of the test. We are sure that you would rather reach the benchmarks than as soon as possible.
We used the Gigabyte X570 Aorus master motherboard to test the Ryzen processors, while the 8th and 9th generation Intel Core processors on the Gigabyte Z390 Aorus Ultra and the new 10th generation core processors on the Asus ROG Maximus XII Extreme have been tested. All configurations were tested with an RTX 2080 Ti, 32 GB DDR4-3200 CL14 memory and a 360 mm all-in-one liquid cooler Corsair Hydro H150i Pro.
We tested the 10900K in two configurations: One is a standard configuration with a loaded XMP profile, with the multi-core expansion setting disabled by default, as it should be. The second result, called "MCE", offers an activated multi-core improvement. This is essentially an overclocked configuration, such as enabling PBO on an AMD Ryzen processor.
During the tests with the Asus ROG Maximus XII Extreme and MCE enabled, the card maintained an all-core clock frequency of 4.9 GHz, but when the MCE was disabled, the frequency dropped to 4.3 GHz. Depending on the workload and duration, Intel advertises an all-core turbo with up to 4.9 GHz. Asus therefore seems to comply with this specification.
After completing all the tests, we started playing around with other motherboards from other manufacturers and found a different boost behavior. When using the MSI Z490 Tomahawk, all cores are executed with 4.9 GHz by default, regardless of the duration. When the MSI version of MCE was activated, which is referred to as "Enhanced Turbo", the all-core frequency of the 10900K was 5.1 GHz. Basically, MSI allows the Tomahawk to run indefinitely in the state of maximum performance, while Asus follows the Intel specification and reduces the frequency after a short time.
This means that the MCE results shown in this test will be very similar to what you will see immediately with the Tomahawk and some other motherboards. We just wanted to point out that the results will likely differ from one review to the next, with each board manufacturer interpreting the Intel specification in its own way. We also used the Asus motherboard for all of our tests as it was provided by Intel for this test. With all of this, let's get out of the way and go over the results.
First we have Cinebench R20 and I think the results shown here will be pretty meaningful for the raw multi-core performance. In its standard configuration, the 10900K scored 6101 points, 25% faster than the 3700X and 9900K. By activating MCE, the performance could be increased by a further 5%, which is quite typical for this setting.
Compared to the 3900X, the 10900K was 19% slower, which means it is positioned between the 8 and 12 core Ryzen processors, and while this contains little news, it is a difficult one for Intel's new 10-core part Place to do this 3900X is regularly available for just over $ 400, making it cheaper and faster. Before we make any serious conclusions, let's look at a lot more data.
Thanks to this 5.3 GHz clock frequency, the 10900K is very strong in single-core performance. This allowed him to score an impressive 549 points, making it 3% faster than the 3950X and 6% faster than the 3900X. This is also an 8% improvement over the 9900K, which tops at 5 GHz.
If we move on to the 7-Zip compression test, we see that the 10900K is again between the 3700X and the 3900X, although it is closer to the 8-core part in terms of performance. Here the 10900K was only 16% faster than the 3700X and thus 17% slower than the 3900X.
The margins are also similar for decompression performance. Here the 10900K was 12% faster than the 3700X, but 25% slower than the 3900X. So again, it's not a great place for Intel's new flagship desktop offering.
The Ryzen processors still largely dominate AES encryption performance, and the SiSoft Sandra benchmark is one of the more balanced tests. Still, the 3900X was 55% faster than the 10900K.
In the Blender Open Data benchmark, the 10900K is only 16% faster than the 9900K and thus 20% slower than the 3900X. Enabling MCE, however, significantly increases performance in this test as the all-core clock speed increases from 4.3 GHz to 4.9 GHz. However, as you'll see later in the test, this has a significant impact on power consumption.
In the V-Ray benchmark, the 10900K offers a performance increase of 27% compared to the 9900K and was therefore 30% faster than the 3700X, but still 13% slower than the 3900X. Activating MCE increased performance by a further 7%. However, you can also use PBO to increase the 3900X by a single digit.
The 10900K does quite well in the Corona benchmark and could keep up with the 3900X with activated MCE. Share even though it was 10% slower.
Here's a look at the performance of code compilation. Here the 10900K is again between the 3900X and the 3700X. Basically, it was 18% faster than the 3700X, but still 20% slower than the 3900X, and that seems to get the 10900K to the point when it comes to productivity performance.
When we move to video editing performance with DaVinci Resolve Studio 16, we see that the 10900K can hardly beat the Ryzen 7 3700X and increases performance by 1.5%. That meant the 3900X was 9% faster, by no means a big margin, but the slightly cheaper Ryzen processor was faster in this test.
The 10900K was more competitive in Premiere Pro as it outperformed the 3700X by 14%, which was only 5% slower than the 3900X. So a decent result and certainly a big performance increase from 9900K.
Photoshop benefits from strong single-core performance. As a result, the 10900K performed very well and even outperformed the 3950X by 7%. So if you're looking for the latest performance in Photoshop, the 10900K is the way to go.
The 10900K was also strong in After Effects and matched the 3900X. Although it wasn't an absolute win, it was just as quick.
Here's a look at power consumption. As you can see when complying with the 125 TDP specification, the 10900K consumes power because the overall system utilization is just below that of the Ryzen 7 2700X. So that's a pretty good result. We should note that as soon as the processor reduces all cores to 4.3 GHz, this is the continued power consumption before the power consumption looks as we see it when the MCE is activated.
Speaking of which, with MCE enabled, the default configuration for many other Z490 cards, power consumption is skyrocketing, and now we see a little over 300 watts for the overall system.
Time for some gaming benchmarks. First, take a look at Battlefield V, which was tested at 1080p using the ultra-quality preset with a GeForce RTX 2080 Ti graphics card. Here, the 10900K essentially matched the 9900K, which meant it was 8% faster than the 3900X and increased the average frame rate from 156 fps to 168 fps.
As expected, this margin shrinks when we switch to the more GPU-bound 1440p resolution, and now the 10900K was only 5% faster than the 3900X.
Far Cry New Dawn shows a worst case scenario for Ryzen and here the 10900K is 18% faster than the 3900X, which is a significant margin. Moving an average of 134 fps versus 114 fps is undoubtedly more desirable for players with a high refresh rate.
The margin remains largely the same at 1440p, since we are largely tied to the CPU. As a result, the 10900K was still 16% faster than the 3900X.
We see fairly typical margins in the new Gears Tactics game. Here the 10900K was 8% faster than the 3900X when comparing the average frame rate and 11% faster when comparing the 1% low results.
Margins are pretty close at 1440p, and here the 10900K was 6% faster than the 3900X when comparing the average frame rate, but interestingly enough, we see no difference for the 1% low result, resulting in a very similar gaming experience.
At Rainbow Six Siege we see a 3% increase in performance for the 10900K over the 9900K with a 5% improvement in 1% low performance. Not exactly stunning, but as Intel promised, the 10900K is now the fastest gaming processor in the world. It was also only 6% faster than the 3900X, but since both are pushing well over 200 fps at any given time, I'm not sure how important this margin is at all.
Once again, 1440p helps bring everything together, and now the 10900K is only 3% faster than the 3900X or 2% when comparing the 1% low results.
Ghost Recon Breakpoint is not a particularly CPU-demanding title. The 10900K was only 5% faster than the 3900X in this game and we would say that this is a fairly accurate representation of modern game performance.
The edge actually opens up a bit at 1440p, now the 10900K is 7% faster than the 3900X, not something to get excited about, but you can enjoy a few extra frames.
Shadow of the Tomb Raider is very CPU demanding, especially the open world area that we use for testing. With that in mind, these results weren't recorded with the built-in benchmark, which we don't think is that demanding and not a good way to test the CPU performance in this title.
We use OCAT to measure in-game performance. Here we see very strong 10900K results when we look at the 1% low data. It was 22% faster than the 9900K and 14% faster than the 9700K, which seems to do better when Hyper-Threading is turned off, although this is not uncommon as we see it quite often.
If you switch to 1440p, the high-end CPUs have a high GPU bottleneck. As a result, the 3900X, 3950X and 10900K all deliver the same 99 fps and 91 fps 1% low on average.
Finally, we have Red Dead Redemption 2 with reduced quality settings and yet we still encounter a strong GPU bottleneck with the high-end CPUs, which limits the performance to a little more than 100 fps on average.
The same goes for 1440p, although the slightly higher single-core performance of the 10900K seems to offer a small advantage, although we only achieve a 2% increase over the 9900K.
Let's look at the average of the seven games. Given what we've just seen, it's no surprise to you that the 10900K is 7% faster on average than the 3900X. At 1080p, this isn't a lot of leeway when using an RTX 2080 Ti, but it means that Intel is keeping the crown of gaming performance.
Operating temperatures were recorded during our Blender stress test with the Corsair Hydro H150i Pro installed. The Core i9-10900K peaked at just 63c, which is lower than expected, but the power consumption is pretty good overall with a case performance of 125 watts. The voltage optimization and a strict binning process are due here, since all 10 cores clocked at 4.3 GHz only required 1.03 V.
However, as soon as you activate MCE and release this additional power, the operating temperature rises to 84 ° C, which in turn is lower than expected when the power of the package reaches 200 watts, and here the CPU was powered by 1.172 V.
What we have learned
After taking a close look at how the Core i9-10900K behaves in applications and games and how it naturally beats the rival AMD Ryzen processors. The question now is, should you buy it?
Let's start by answering this question based on the performance we've seen so far. When gaming, the new 10-core Intel processor was up to 18% faster than the 3900X, but ~ 7% faster on average, and frankly, it's hard to tell the difference. In fact, it is simply impossible for the vast majority of games. Not to mention that the 7% difference at 1080p was observed with an RTX 2080 Ti. When we compare the 1440p data, the 10900K is only 5% faster than the 3900X when comparing the average frame rate and 1% low data.
In productivity applications, we found that the 10900K was up to 35% slower than the 3900X, and there have been a number of cases where it was 15% or more slower. In the rare case where the Core i9 processor was actually faster, we also looked at single-digit gains. For example, you will almost certainly not see a 9% increase in performance in Photoshop, but you will certainly notice a 15% or more decrease in compression, code compilation, or rendering / encoding performance.
It's also about performance per watt. The 10900K is very power-hungry due to unfortunate circumstances in which Intel stuck to the 14 nm process for significantly longer than intended. For the most part, we don't see this as too big a problem. Using a decent cooler keeps the 10-core Comet Lake CPU cool enough, but still uses a lot more power and in applications like Blender, it does so with significantly lower performance.
We believe the Core i9-10900K is a hard sell in terms of performance and has been scaled down to become a niche product. Who should want to buy the 10900K? Maybe someone who wants ultimate gaming performance and better video playback than a 10700K, but for some reason doesn't want to sacrifice gaming performance in favor of even better video playback at a lower price with the 3900X.
With the Ryzen 9 3900X for around $ 410 – certainly not a coincidence, AMD tries to put Intel under pressure here – the CPU has been available for about this price since March. The 10900K earns a serious premium of $ 530, the current pre-order price now. In other markets like Australia, we see prices that are ~ 30% higher than the 3900X. This is a massive price premium that is simply not worth paying. If the 10900K arrives at $ 410, we could certainly make an argument for buying via AMD's 3900X, especially if you are just playing.
However, the biggest problem with the Core i9-10900K may not be performance or price, but availability. There has been a lot of speculation recently that the release of the 10th generation core series is closer to a paper launch than Intel would like to admit. We strongly believe that this is actually the case when pre-orders were opened a few weeks ago, retailers have very low inventory levels and demand was not insane, but the supply has been greatly exceeded so that retailers can only receive orders that cannot be filled . In other words, the chances of getting a Core i9-10900K in the short term are extremely slim. As far as we know, Intel is currently not promising delivery to retailers. If the offer is so limited, it will lead to a price reduction and ultimately lead to buyers choosing Ryzen instead.
It's also crazy to believe that six months have passed since the Ryzen 9 3950X was released and Intel still has no answer, so AMD can get the highest price on the desktop. Just two years ago, Intel charged $ 1,000 for its full 10-core desktop part, the Core i9-9900X. Today you get a slightly faster chip for half of it, more evidence that competition is a good thing.
In terms of motherboard pricing, there appear to be some decent Z490 options for under $ 200, which seems reasonable and at least comparable to AMD's X570 range.
Later this week, we'll look at the Core i7-10700K, followed by the Core i5-10600K, and then we hope to be able to get our hands on some Core i3 models, though there's no word on availability there either. That will do it for our first look at the 10900K and tell us what you think about this new part of Intel.
- AMD Ryzen 9 3900X at Amazon
- AMD Ryzen 9 3950X at Amazon
- Intel Core i9-10900K at Amazon (soon?)
- Intel Core i9-9900K at Amazon
- AMD Ryzen 7 3700X on Amazon
- AMD Ryzen 5 3600X on Amazon
- GeForce RTX 2070 Super on Amazon
- GeForce RTX 2060 Super on Amazon