Rocket Lake is finally here and today we are testing the new Core i5-11600K. That's right, for this generation we're skipping a first look at the Core i9 and even the Core i7 part and instead moving straight to the cheaper Core i5. There are several reasons for this. First, the 8-core Core i7 and Core i9 processors are essentially identical, and initial reviews of these chips were done weeks before the embargo.
For those of you interested, we'll have another review of the Core i9-11900K in a few days' time, but with an MSRP of $ 540, you can probably guess my thoughts on that. The $ 400 11700K is a little tastier, especially when it's widely available. This makes the 11600K the third unlocked 11th generation Rocket Lake chip announced so far.
Like the 10600K, the 11600K is a 6-core / 12-thread CPU that includes a 12MB L3 cache and 125-watt TDP along with the same MSRP of $ 262. The new 11th generation model clocks 100 MHz higher and increases to 4.9 GHz. Due to the AVX-512 support, the base clock was reduced from 4.1 GHz to 3.9 GHz.
It is worth noting that the 11th generation architecture is completely different from that of the 10th generation. Rocket Lake is a hybrid architecture that is the result of two different backport technologies. It's a little tricky, but Intel took the 10nm Sunny Cove cores from the Ice Lake microarchitecture and redesigned it for 14nm to create something completely new they call Cypress Cove.
|Core i5-11600K||Core i5-10600K||Core i5-10600KF||Ryzen 5 5600X|
|Price (RRP)||$ 262||$ 237||$ 299|
|Current retail||$ 270||$ 224||$ 200||$ 350|
|Cores / threads||6/12|
|Base frequency||3.9 GHz||4.1 GHz||3.7 GHz|
|Max turbo||4.9 GHz||4.8 GHz||4.6 GHz|
|L3 cache||12 MB||32 MB|
|iGPU model||UHD graphics 750||UHD graphics 630||N / A|
|TDP||125 watts||65 watts|
But it's more than that. Intel also took the Xe graphics from 10nm Tiger Lake and ported them back to this new hybrid 14nm design, although the name stays the same. It's a bit of a mess, and we have to wonder why they didn't just scrap this hybrid architecture and instead focus on preparing Alder Lake.
"Alder Lake" is Intel's 12th generation core series and is scheduled for release later this year. It is expected to use their refined and highly anticipated 10nm process on a new LGA 1700 socket and have DDR5 memory support.
In the not-too-distant future releases aside, let's take a look at the Core i5-11600K and see if it's a worthy replacement for the now very affordable 10600K, while also beating the overpriced Ryzen 5 5600X.
We used the Gigabyte Aorus Z590 Master with BIOS version F5a to test the Intel CPUs. The board was configured with 32 GB DDR4-3200 CL14 dual-rank, dual-channel memory and the cooling was provided by the Corsair iCUE H150i Elite Capellix White. This is essentially a 360mm AIO. All hardware was installed in the Corsair 5000D Airflow case, which is powered by the RM850x power supply. The Ryzen test system has the black version of the same cooler, memory and power supply. The key change is the Gigabyte X570 Aorus master motherboard.
This test only looks at results for Intel CPUs that are not limited to performance, therefore no TDP-limited tests. This is the way we usually test Intel CPUs, as it is also how most Z490 and Z590 cards are out of the box. So while we're using the standard clock multiplier tables, none of the Intel CPUs adhere to performance limits.
Eventually, all configurations were tested with an RTX 2080 Ti for productivity benchmarks, but for gaming, all benchmarks were run with the newer and faster GeForce RTX 3090. Let's get to the results.
Starting with Cinebench R20, we see an impressive performance increase of 20% for the 11600K compared to the previous generation. While that wasn't enough to outperform AMD's 6-core offering, it did bring Intel up to speed. The 5600X is only 3% faster, which is a negligible margin.
Basically, AMD's simultaneous multithreading technology still seems to be a bit more powerful, as Intel can keep up with AMD in terms of single-core performance, as both the 11600K and 5600X achieve a score of around 600 points.
The gains on the 7zip compression benchmark aren't all that impressive, although we still expect a solid 14% increase over the 10600K. For Intel, this meant that the new 11600K outperformed the 5600X by 12% in this test.
We're also seeing a 14% increase over the 10600K for decompression work, and again, Intel's SMT implementation is not as effective as AMD's, resulting in a 19% deficit over the 5600X.
The 10th Gen Intel Core series was particularly weak for AES-256 encryption, but here we see a huge 57% performance increase for the 11900K, which is 6% ahead of the 5600X. An incredible increase in performance and this is the first benchmark we've seen where Intel is now beating AMD.
The 11600K took another nice step forward, this time in the Blender Open Data benchmark, where it matched the 5600X and offered a 15% increase in performance over the 10600K.
Rendering performance in V-Ray was also significantly improved, with the 11600K performing 22% more than the 10600K, putting it back within range of the 5600X.
We're seeing a 14% increase in performance in the Corona benchmark and that was enough to get within 5% of the 5600X. Here the AMD processor was 4% faster, which is a pretty insignificant margin.
Next we have the compilation performance of Chromium code and again the 11600K delivers a 14% performance improvement over the 10600K, which is hot on the heels of the 5600X. Again, the AMD processor was only 2.5% faster, which is a negligible margin.
Even in the DaVinci Resolve Studio 16 benchmark, we see a 10% improvement in the 11th Gen Core i5 over its predecessor, and that's basically the same level of performance that we got from the Ryzen 5 5600X.
Premiere Pro's profits are even bigger. Here the 11600K outperformed the 10600K by 16%, which in turn meant it could deliver 5600X-like performance.
The performance increase in Photoshop for the 11600K was enormous, outperforming the 10600K by 21%. That improvement was enough to see that the new 11th generation part only outperforms the 5600X, although this is really a comparable performance.
The performance margins found when testing with After Effects are also very similar. Here the 11600K offered a 19% increase in performance compared to the 10th generation model and was thus roughly on par with the 5600X.
This is where things go a bit wrong for Intel and their 10nm issues, as they have stuck to the tried and true 14nm process for seven generations spanning six years. While the Core i5-11600K can keep up with the Ryzen 5 5600X in almost all of our productivity benchmarks, it increased overall system utilization by 41% to achieve this. The following results measure the average power consumption across the entire Blender Open Data benchmark.
To move on to the gaming benchmarks, we have swapped the RTX 2080 Ti for the more powerful RTX 3090. Starting with Watch Dogs Legion, we see a slight 6% increase in performance over the 10600K, which puts the 11600K on par with the 5600X.
Interestingly, the frame rates stay the same in F1 2020 as the 11600K is the same as the 10600K and the 5600X stays 10% faster, or 13% faster if we compare the 1% low data.
The new 11th Gen Core i5 CPU still hit over 240 fps on average, so margins probably don't matter, but it wasn't nice to see the 11600K lag behind the 5600X in terms of productivity have not particularly seen benchmarks.
Next up we have Horizon Zero Dawn and here we see comparable performance to the Ryzen 5 5600X. The 1% low performance is the same, while the AMD CPU was only 3% faster on average. The 11600K is also 3% faster than the 10600K, which means the three CPUs in this title deliver similar performance.
Intel was already slightly ahead of AMD in Borderlands 3, and the 11600K manages to increase that margin and increase performance by 4% over the 5600X. Not exactly earth-shattering, but it was good to see Intel making some profits here.
The frame rates between 11600K and 5600X are identical with Death Stranding. While this might not seem overly impressive at first glance, you must realize that the 11600K is 18% faster than the 10600K, which is an impressive increase in performance for the new Intel processor.
We expect a 7% increase in performance for the 11600K over the 10600K in Hitman 2, or 10% when we compare the 1% lower performance. That's a decent boost, but it wasn't enough to catch the 5600X, which is still leading by an 8% margin.
Unfortunately, performance in Star Wars Squadrons dropped as the 11600K dropped from 259 fps to 249 fps with the 10600K, which is a small 4% drop in performance. The 5600X already had a big lead over the 10600K in this title. The result is that the AMD processor is 21% faster than the 11600K, which is a significant margin, albeit downplayed by the extreme frame rates.
Serious Sam 4 shows little to no improvement in performance for the 11600K compared to the 10600K. The 5600X stays 14% faster in this title, which is a significant advantage.
The results of Rainbow Six Siege are a bit inconsistent. The 11600K improved performance by 1% by 6%, but then we saw a 2% reduction in the average frame rate. The end result was that the 5600X performed 14% faster and reached no less than 506 fps.
The village we use for testing in Shadow of the Tomb Raider is very CPU demanding and here the 11600K delivers a huge 11% increase in performance over the 10600K which is great to see. It's still not the fastest 6-core processor in the game as the 5600X was another 8% faster, averaging 159 fps.
Average gaming performance
If we look at the average of 10 games, we see that Intel couldn't recapture the gaming crown, at least not for 6-core CPUs. Overall, we're seeing a small 5% improvement over the 10600K, and while that is to be welcomed, AMD is up 7% this time around. There's not much more to say here as we suspect the 5600X's massive 32MB L3 cache is just too much for the 11600K.
When it came time to overclock, we were only able to stabilize our 11600K sample at 4.9 GHz. It was possible to start Windows at 5 GHz, but even with an uncomfortable voltage, we couldn't stabilize the overclocking.
At 4.9GHz it was completely stable at 1.35V, and that 6.5% increase in all-core frequency increased the Cinebench R20 multithreaded score by only 3%.
The 10600K, on the other hand, scaled far better with its 13% overclocking from a core of 4.5 GHz to 5.1 GHz, as this increased the score by 12%. This is interesting for those looking to overclock these processors. While the 11600K was 20% faster than the 10600K, it is only 10% faster when both CPUs are overclocked to the maximum.
The results of Rainbow Six Siege are also interesting. The 11600K saw a 7% increase after overclocking, while the 10600K saw less than one percent. Now you may be thinking that this is not possible, and I admit it certainly doesn't look right, but these results are based on a three run average and we did a triple check.
So there seems to be a bottleneck with the 10600K that is not frequency related. We've consistently seen increases from the 1% lows, which overclocking has improved by around 8%.
Finally, let's take a look at the overclocking performance with Watch Dogs Legion as we saw it with Cinebench R20. Little can be gained by overclocking the 11600K. We also only see a 3% overclock for the 10600K.
The value proposition
The new Core i5-11600K CPU is usually between 10% and 20% faster than the 10600K for productivity tasks, which corresponds to Intel's requirement for an IPC increase of 20%. That means the new part can compete with the Ryzen 5 5600X in most scenarios, so Intel has done well in terms of productivity. They are nowhere near AMD in terms of efficiency, and of course that's down to the process node.
When it comes to games, AMD is still ahead of the game. The Ryzen 5 5600X is about 7% faster than the 11600K, based on our 10-game example. Intel made a small jump of 5% over the 10600K here, although we saw significant gains with some titles like Death Stranding, with some others there were hardly any differences.
It's worth noting that gaming benchmarks were run with an RTX 3090 at 1080p. This is used to measure pure CPU performance, although under more realistic conditions like playing games at 1440p or higher, the 7% margin of the 5600X over the 11600K is significantly reduced and in most cases eliminated entirely. While the Ryzen 5600X is the better gaming CPU, it's not worth paying a high premium for the difference, as was the case with the 9th and 10th generation Core series versus AMD's Zen 2 series .
At the time of writing, we could determine that the Ryzen 5 5600X was selling for around $ 365 on Newegg and Amazon, which is $ 65 above MSRP. The Core i5-11600K is now listed on Amazon for $ 270, which is 26% cheaper than the 5600X. Honestly, the Ryzen part isn't worth that price premium.
If both CPUs were available at their MSRP, the 5600X would be the obvious choice because it's faster, more efficient, and supported by a wider range of motherboards. But at today's prices we would get the 11600K. No, we would forget about both the 5600X and 11600K and only get the outgoing 10600K, or better yet, the 10600KF.
The Core i5-10600K is priced at $ 224, and the 10600KF (no graphics) model is even more attractive at $ 200. That 35% savings could then flow into something else, especially if your focus is on gaming. If you need a CPU upgrade today and are on a budget between $ 200 and $ 300, the Core i5-10600KF is a good CPU for the money.
If you're looking for an inexpensive motherboard to do this, check out our top 5 best Z490 motherboards. The MSI Pro Z490-A Pro was a good option and is still available for around $ 170. There are also a couple of Z590 motherboards for under $ 200 that may be better, although we suspect a Z490 board would be better off with a 10th generation processor.