Most likely, you’ve already read our review of the senior member of the Rocket Lake family – the Core i9-11900K processor. And most likely, this novelty made the same impression on you as on us: with all due respect to the work invested in the implementation of this project, I don’t want to use such a processor, to put it mildly. Of course, Rocket Lake is a big step for Intel, as it has updated the microarchitecture in a desktop CPU for the first time in six years, and, moreover, even achieved a tangible increase in IPC. But the problem is that, judging by the results of testing the Core i9-11900K, this microarchitecture clearly should not have been published using 14nm design standards. However, Intel decided that the old process technology could be applied one more time, and because of this we got a chip with monstrous heat dissipation, which immediately gave rise to a host of various problems.

But everything that has been said above is not yet an epicrisis, but only a preliminary diagnosis. The fact is that it is not entirely legitimate to judge the entire Rocket Lake family by the older model alone. The peculiarity of the Core i9-11900K lies in the fact that Intel has set a super task for it – by all means return the title of the best processor for games. And in the name of her, the characteristics of this model were twisted with an obvious bias towards overestimating clock frequencies, which dealt a serious blow to operating temperatures and power consumption. In other words, the Core i9-11900K is a severely overclocked processor at the factory, with all the ensuing consequences.

Fortunately, along with the Core i9-11900K in the Rocket Lake family, there are other models that are more balanced in their characteristics. There are such ones even among eight cores: they are one step below the flagship in the Intel hierarchy and belong to the Core i7 series. It is noteworthy that the CPUs of this series repeat most of the characteristics of the Core i9-11900K, except perhaps for the exorbitant frequencies, and, it seems, should not be as fiery chips as their older brother. For this review, we took the overclocking processor like the Core i7-11700K and hope it can correct the mixed first impression of Rocket Lake.

Interest in the Core i7-11700K is supported by its price. Despite the fact that it loses only a little to the Core i9-11900K in clock speeds, its price is set at $ 399 versus $ 539 (according to the official price list)  the older model… Therefore, consumers who are not allergic to Intel products and are guided by rational considerations are more likely to choose the Core i7-11700K over the Core i9-11900K. The only question is how much the Core i7-11700K, due to the rebalancing of the characteristics, managed to get rid of the disadvantages of the flagship Rocket Lake, and whether all the advantages of the new processor design are not lost from this. We will try to give the answer to this very question in this article.

⇡ # More about Core i7-11700K

The Core i7-11700K is an eight-core Rocket Lake generation processor, much like the flagship. They are not only  are similarin terms of the number of cores and the amount of L3 cache, but they also do not differ in overclocking functions: Core i7-11700K has the same set of unlocked multipliers and a dual-mode memory controller, with which DDR4 SDRAM can work both synchronously and at doubled frequency. The differences are only in operating frequencies and in the number of supported versions of turbo modes: in the end, the Core i7-11700K lacks Thermal Velocity Boost and Adaptive Boost, which leads to its lagging behind in maximum frequencies from its older brother by 300-500 MHz.

In this case, there are two curious nuances. First, the frequency reduction does not translate into thermal packet changes. The calculated heat dissipation of the Core i7-11700K is the same 125 W as that of the Core i9-11900K. The second point is the unexpected 100 MHz superiority of the Core i7-11700K over the Core i9-11900K at the base frequency. This can be read in such a way that for the Core i9 and Core i7 series, the manufacturer chooses crystals of different quality. It seems that crystals with lower leakage currents are used for the Core i7-11700K, which means that, all other things being equal, the power consumption of such processors may indeed be lower.

All this is reflected in the table, in which, together with the indicators of Core i7-11700K and Core i9-11900K, for clarity, the characteristics of the older eight-core Intel processor of the last generation are also given.

	Core i9-11900K	Core i7-11700K	Core i7-10700K
Platform	LGA1200	LGA1200	LGA1200
Microarchitecture	Cypress cove	Cypress cove	Skylake
Technological process, mm	14	14	14
Kernels / threads	8/16	8/16	8/16
Frequency (nominal / turbo), GHz	3,5-5,3	3,6-5,0	3,8-5,1
Full load frequency, GHz	5,1	4,6	4,7
L2 cache, KB	× 8 512	× 8 512	× 8 256
L3 cache, MB	16	16	16
AVX-512	There is	There is	No
TDP, W	125	125	125
Memory	DDR4-3200	DDR4-3200	DDR4-2933
PCIe Lines	20 × Gen 4	20 × Gen 4	16 × Gen 3
Built-in graphics	UHD 750	UHD 750	UHD 630
Price	$ 539	$ 399	$ 387

We have already paid attention to how much the size of the Rocket Lake crystal has increased compared to its predecessors, while the technological process remains unchanged. Even compared to the ten-core version of Comet Lake, the die area has grown by about a third and is now 276 mm²… Therefore, do not be surprised that in terms of clock speeds, the Core i7-11700K is slightly inferior to the eight-core of the previous generation, the Core i7-10700K. Fortunately, the massive gain in IPC provided by the Cypress Cove microarchitecture implemented in Rocket Lake should compensate for this flaw.

At the same time, the Core i7-11700K and i7-10700K differ in the passport consumption limits of PL1 and PL2. For the new processor, they are the same as for the Core i9-11900K, that is, before it enters the declared thermal package, it is allowed to consume up to 56 watts of electricity for 251 seconds. With Core i7-10700K, the maximum consumption limit was lower – it was limited to 229 watts. However, we’d better not rely on the manufacturer’s declarations, but simply measure everything.

The graph below shows the practical consumption of the Core i7-11700K in Cinebench R23 under load on various numbers of threads. At the same time, the PL1 and PL2 limits are disabled, that is, we estimate how much power the Core i7-11700K may need if it is not limited by anything.

Based on consumption in Cinebench R23, it turns out that the limit  PL2251 W is installed for the Core i7-11700K with a large margin. In reality, the processor consumes less than 190 W even with the maximum load on all cores, and this is much lower than the consumption of the Core i9-11900K. There, we recall, at full load, the values ​​ u260b u270bare recorded at the level of XNUMX-XNUMX W, and after all, we are talking in fact about the same eight-core processor.

However, the Core i125-1K naturally does not fit into the 7W frame set by the TDP and PL11700 limit. This means that after some time (limited from the top by an interval of 56 seconds) it will reset its frequency. The magnitude of this slowdown is estimated in the following graph, which shows the frequency curve for the PL1 and PL2 limits, built on the basis of real measurements when rendering in Cinebench R23 with a load on a different number of cores.

Differences in frequencies with different consumption limits begin with an increase in the number of active streams above five. At the same time, at maximum load, the frequency of the Core i7-11700K, limited by the 125 W limit, is reduced to 4,0-4,1 GHz, which is approximately 500-600 MHz lower than the maximum possible frequency in this state.

True, all these measurements have more theoretical than practical meaning. In reality, motherboards based on the older Z490 and Z590 chipsets ignore the PL1 and PL2 limits and always move the processor to the highest possible frequency, which depends solely on the number of cores loaded with work. In the case of the Core i7-11700K, this means a frequency of 5,0 GHz with a load of 1-3 cores, 4,9 GHz – with a load of 4 cores, 4,7 GHz – with a load of 5-6 cores and 4,6 GHz – at a higher load.

Along with the clock frequency of the computing cores, the Core i7-11700K also changes the frequency of the extra-core part (Uncore) – the ring bus and L3 cache. In most cases, it is 800 MHz lower than the processor frequency, that is, for example, it is 3,8 GHz at full load. Accordingly, in terms of this characteristic, the Core i7-11700K is slightly inferior to the Core i9-11900K, and even more so to the Core i7-10700K, which belongs to the previous generation. However, nothing terrible happened: even with a reduced Uncore frequency, the considered representative of the Rocket Lake family wins in terms of inter-core delays over its predecessor, the Core i7-10700K. The explanation is simple – in eight-core Rocket Lake, the ring bus is trite to be shorter, because Comet Lake processors are designed with a ten-core design in mind.

However, the lower Uncore frequency still affects the speed of memory and L3 cache. It’s easy enough to compare the practical latencies measured between the Core i7-11700K and the Core i7-10700K by the AIDA64 Cachemem benchmark.

Core i7-11700K, DDR4-3600 Gear 1

Core i7-10700K, DDR4-3600

The RAM and L7 cache of the Core i11700-3K work with slightly higher latencies, and the LXNUMX cache also loses in terms of practical bandwidth.

But here you can see some positive aspects, first of all, the fact that the memory controller in the Core i7-11700K works the same as in the older Core i9-11900K. For example, in the case of the DDR4-3600 SDRAM with which we conducted tests, the Core i7-11700K controller set Gear 1’s fast synchronous mode, and everything worked perfectly stable. That is, despite the fact that the official specification for the Core i7-11700K only claims DDR4-3200 support in the Gear 2 slow mode, in reality its memory controller does not differ from the Core i9-11900K memory controller, and Gear 1’s synchronous mode with high-speed memory (with a frequency of at least up to DDR4-3733) works without problems.

⇡ # Consumption and temperatures: is everything terrible?

After getting acquainted with the Core i9-11900K about the thermal and power characteristics of the Rocket Lake processors, an extremely negative impression was formed. But the older eight-core processor was clearly brought up to the maximum frequencies, and the model of the Core i7 series, where Intel did not pursue the highest possible performance, it seems, should not be so gluttonous and hot. However, this hypothesis needs more detailed testing.

The following graphs compare the instantaneous power consumption of Core i7-11700K, Core i7-10700K, and Ryzen 7 5800X when passing three different tests: when rendering with a processor in Blender 2.91, in the Prime95 30.3 stress test with AVX2 instructions enabled, and in the game Horizon Zero. Dawn.

Whatever one may say, Rocket Lake turns out to be gluttonous from birth. How much the Core i9-11900K got hot was only partly due to its factory overclocking via Adaptive Boost technology. A noticeable contribution is made by the Cypress Cove microarchitecture itself, which is clearly seen in the consumption of the Core i7-11700K, which under load does not operate at 5,1 GHz, as its older brother, but at a quieter 4,6 GHz. Nevertheless, both in rendering in Blender and in the AVX2-intensive Prime95 test, the new Core i7-11700K consumes more than the previous generation Intel eight-core processor by a weighty 40-45W.

The presence on the presented graphs of consumption indicators for the eight-core Ryzen 7 5800X further aggravates the bewilderment of the Core i7-11700K’s appetites. In Blender, the current AMD processor requires one and a half, and in Prime95, half as much power as the Rocket Lake processor under review.

A slightly different situation is observed only in the game, where the load on the processor is fundamentally lighter. But even there, against the background of an average consumption of Ryzen 7 5800X and Core i7-10700K of about 90-95W, the new Core i7-11700K requires all 120W.

Together with the consumption, we measured the temperatures of the processors when they passed the same tests. This measurement is of special interest, because the temperature regime depends not only on the heat dissipation of the CPU, but also on the efficiency of the heat sink, which is directly related both to the quality of the thermal interface under the processor cover and to the area of ​​the semiconductor crystal itself. For example, it is easier to remove heat from the giant Rocket Lake crystal than from the smaller Comet Lake semiconductor crystal. And cooling the CCD chiplet of the Ryzen 7 5800X processor, which is three and a half times smaller than Rocket Lake, is a whole problem. We used the same liquid cooling system to cool all three processors, and the results are as follows.

The undisputed anti-leader in terms of temperatures is the Ryzen 7 5800X. Under heavy loads, its cores confidently reach the maximum temperature level for this processor of 90 degrees. In the game, its temperature regime is better, but the heating still reaches 75 degrees. Against this background, the Core i7-11700K even seems cold: in Blender its temperature only slightly goes beyond the 70-degree mark, and in Horizon Zero Dawn it stays at about 55 degrees. Heating of the Core i7-11700K above 80 degrees can be seen only in the case of AVX2 load in Prime95. In other words, judging by the temperatures, the Core i7-11700K is only slightly hotter than its predecessor, and the level of heat dissipation that is characteristic of this processor can be quite put up with.

⇡ # Overclocking

An important question that largely determines the appeal of the Core i7-11700K is whether it can run at the same clock speeds as the flagship Core i9-11900K. It is assumed that Intel uses silicon with lower leakage currents for the representatives of the Core i7 series, which may mean that such processors may increase their frequency potential worse with increasing voltage. This turned out to be partly true for our Core i7-11700K instance – we managed to achieve stable operation of this CPU only at 5,0 GHz, while the flagship was able to be overclocked to a frequency of 100 MHz higher.

However, the difference is not so great, and in any case, overclockers will be able to get a few extra percent of the performance from the Core i7-11700K, which will bring it as close as possible to the speed of the Core i9-11900K. To overclock Core i7-11700K, we adjusted its power supply in Offset mode, through which the voltage curve was raised by 0,05 V. The Load-Line Calibration function – Vdroop counteraction – was transferred to the Level 4 state. With such settings, the processor did not have problems in stability tests.

Note that when overclocked to 5,0 GHz, stability testing in Prime95 brought the Core i7-11700K temperature to 90 degrees, and the consumption to 280 W. And this is work without AVX instructions. Their execution in Rocket Lake leads to a significant increase in heat dissipation, therefore, in order to avoid overheating, it is necessary to use decreasing corrections to the CPU multiplier.

So, for AVX / AVX2 instructions the frequency had to be limited to 4,8 GHz, and for AVX-512 – 4,6 GHz. But in this case, overheating of the Core i7-11700K was no longer threatened.

As you can see from the screenshot above, the temperature of the overclocked Core i7-11700K with AVX2 load in Prime95 is about 100-105 degrees, which can be considered acceptable for a heavy stress test.

At the same time, do not forget for a second that we are talking about overclocking an initially very hot processor, made using a 14-nm process technology. Its consumption and heat dissipation during overclocking can easily exceed the 300 W line. And in order to remove this amount of heat, a high-performance cooling system is required. For example, in our overclocking experiments, we used the EKWB LSS, and we should not expect that such overclocking could be achieved with an air cooler.

⇡ # Description of the test system and testing methodology

If we proceed from Intel’s pricing policy, then the Core i7-11700K is logical to be considered a more modern replacement for the eight-core of the previous generation, the Core i7-10700K. The main competitor to the Core i7-11700K from AMD, apparently, will be the modern eight-core Ryzen 7 5800X, although it is $ 50 more expensive. But in fairness, it should be noted that the Ryzen 7 11700XT, which belongs to the previous generation, has a price closer to the Core i7-3800K, so it should not be written off either.

The four processors mentioned in the paragraph above have become the backbone of today’s testing, but we decided not to limit ourselves to them only. Processors of the corresponding generations, which are one step higher, were also involved in the test. In particular, the new flagship Core i9-11900K, whose participation will make it possible to conclude what a performance loss will result in the 140-dollar savings when installed in a PC, not it, but the Core i7-11700K.

Thus, the test system includes the following components:

• Processors:
  • AMD Ryzen 9 5900X (Vermeer, 12 cores + SMT, 3,7-4,8 GHz, 64 MB L3);
  • AMD Ryzen 7 5800X (Vermeer, 8 cores + SMT, 3,8-4,7 GHz, 32 MB L3);
  • AMD Ryzen 9 3900XT (Matisse, 12 cores + SMT, 3,8-4,7 GHz, 64 MB L3);
  • AMD Ryzen 7 3800XT (Matisse, 8 cores + SMT, 3,8-4,7 GHz, 32 MB L3);
  • Intel Core i9-11900K (Rocket Lake, 8 cores + HT, 3,5-5,3 GHz, 16 MB L3);
  • Intel Core i9-10900K (Comet Lake, 10 cores + HT, 3,7-5,3 GHz, 20 MB L3);
  • Intel Core i7-11700K (Rocket Lake, 8 cores + HT, 3,6-5,0 GHz, 16 MB L3);
  • Intel Core i7-10700K (Comet Lake, 8 cores + HT, 3,8-5,1 GHz, 16 MB L3).
• CPU cooler: custom LSS EKWB.
• Motherboards:
  • ASUS ROG Crosshair VIII Hero (Socket AM4, AMD X570);
  • ASUS ROG Maximus XIII Hero (Wi-Fi) (LGA1200, Intel Z590).
• Memory: 2×16 GB DDR4-3600 SDRAM, 16-18-18-38 (Crucial Ballistix RGB BL2K16G36C16U4BL).
• Video card: NVIDIA GeForce RTX 3090 Founders Edition (GA102, 1395-1695 / 19500 MHz, 24 GB GDDR6X 384-bit).
• Disk subsystem: Intel SSD 760p 2 TB (SSDPEKKW020T8X1).
• Power supply: Thermaltake Toughpower DPS G RGB 1000W Titanium (80 Plus Titanium, 1000W).

All compared processors were tested with the default settings of the motherboard manufacturers. This means that for Intel platforms, the power consumption limits specified in the specifications are ignored, and instead, the maximum possible frequencies are used in order to obtain maximum performance. The overwhelming majority of users operate processors in this mode, since the activation of heat dissipation and power consumption limits in most cases requires special BIOS settings.

RAM in AMD and Intel systems was configured in DDR4-3600 mode with XMP timings. This also means that Ryzen processors used Synchronous Memory Controller mode and 1800 MHz Infinity Fabric frequency, while the latest generation Core processors used Gear 1 mode and 1800 MHz memory controller frequency.

Testing was performed on Microsoft Windows 10 Pro (20H2) Build 19042.572 using the following set of drivers:

• AMD Chipset Driver 2.13.27.501;
• Intel Chipset Driver 10.1.31.2;
• NVIDIA GeForce 461.40 Driver.

Description of the tools used to measure computational performance:

Complex benchmarks:

• Futuremark PCMark 10 Professional Edition 2.1.2508 – Testing in scenarios Essentials (typical work of the average user: launching applications, surfing the Internet, video conferencing), Productivity (office work with a word processor and spreadsheets), Digital Content Creation (digital content creation: editing photos, nonlinear video editing, rendering and visualization of 3D models).
• 3DMark Professional Edition 2.17.7173 – testing in the Time Spy Extreme 1.0 scene.

Applications:

• 7-zip 19.00 – testing the speed of archiving. The time taken by the archiver to compress a directory with various files with a total volume of 3,1 GB is measured. The LZMA2 algorithm and the maximum compression ratio are used.
• Adobe Photoshop 2021 22.2.0 – performance testing for graphics processing. This measures the average execution time of the Puget Systems Adobe Photoshop CC Benchmark 18.10 test script, which simulates typical processing of a digital camera image.
• Adobe Photoshop Lightroom Classic 10.11 – performance testing when batch processing a series of images in RAW format. The test scenario includes post-processing and JPEG export at 1920 × 1080 resolution and maximum quality of two hundred 16MP RAW images taken with a Fujifilm X-T1 digital camera.
• Adobe Premiere Pro 2020 14.9.0 – performance testing for non-linear video editing. This measures the render time to YouTube 4K of a project containing HDV 2160p30 footage with various effects applied.
• Blender 2.91.2 – testing the speed of the final rendering in one of the popular free packages for creating three-dimensional graphics. The time taken to build the final pavillon_barcelona_v1.2 model from Blender Benchmark is measured.
• Cinebench R23 is the standard benchmark for testing rendering speed in Cinema 4D R23.
• Magix Vegas Pro 18.0 – performance testing for non-linear video editing. This measures the render time to YouTube 4K of a project containing HDV 2160p30 footage with various effects applied.
• Microsoft Visual Studio 2017 (15.9.33) – measuring the compilation time of a large MSVC project – a professional package for creating three-dimensional graphics Blender version 2.79b.
• Stockfish 12 – testing the speed of a popular chess engine. The speed of enumerating options in the position “1q6 / 1r2k1p1 / 4pp1p / 1P1b1P2 / 3Q4 / 7P / 4B1P1 / 2R3K1 w” is measured.
• SVT-AV1 v0.8.6 – testing the speed of video transcoding into the promising AV1 format. To evaluate the performance, the original 1080p @ 50FPS AVC video file is used, which has a bitrate of about 30 Mbps.
• Topaz Video Enhance AI v1.7.1 – performance testing in an AI-based program to improve video detail. The test uses the original video at 640 × 360, which is doubled using the Artemis LQ v7.
• V-Ray 5.00 – testing the performance of the popular rendering system using the standard V-Ray Benchmark Next application.
• VeraCrypt 1.24 – Cryptographic Performance Testing. A benchmark built into the program is used, which uses Kuznyechik-Serpent-Camellia triple encryption.
• x265 3.5 + 8 10bpp – testing the speed of video transcoding to H.265 / HEVC format. For performance evaluation, the original 2160p @ 24FPS AVC video file is used, which has a bitrate of about 42 Mbps.

Games:

• Assassin’s Creed Odyssey. 1920 × 1080 resolution: Graphics Quality = Ultra High. Resolution 3840 × 2160: Graphics Quality = Ultra High.
• Borderlands 3. Resolution 1920 × 1080: Graphics API = DirectX 12, Overall Quality = Badass. 3840×2160 resolution: Graphics API = DirectX 12, Overall Quality = Badass.
• Civilization VI: Gathering Storm. 1920 × 1080 resolution: DirectX 12, MSAA = 4x, Performance Impact = Ultra, Memory Impact = Ultra. 3840×2160 resolution: DirectX 12, MSAA = 4x, Performance Impact = Ultra, Memory Impact = Ultra.
• Crysis Remastered. 1920 × 1080 resolution: Graphics Settings = Very High, RayTracing Quality = Very High, Anti-Aliasing = TSAA. Resolution 3840 × 2160: Graphics Settings = Very High, RayTracing Quality = Very High, Anti-Aliasing = TSAA.
• Cyberpunk 2077. Resolution 1920 × 1080: Quick Preset = Ray Tracing – Ultra. 3840×2160 resolution: Quick Preset = Ray Tracing – Ultra.
• Far Cry New Dawn. 1920 × 1080 resolution: Graphics Quality = Ultra, HD Textures = On, Anti-Aliasing = TAA, Motion Blur = On. 3840 × 2160 resolution: Graphics Quality = Ultra, Anti-Aliasing = Off, Motion Blur = On.
• Hitman 3. Resolution 1920 × 1080: Super Sampling = 1.0, Level of Detail = Ultra, Texture Quality = High, Texture Filter = Anisotropic 16x, SSAO = Ultra, Shadow Quality = Ultra, Mirrors Reflection Quality = High, SSR Quality = High, Variable Rate Shading = Quality. Resolution 3840×2160: Super Sampling = 1.0, Level of Detail = Ultra, Texture Quality = High, Texture Filter = Anisotropic 16x, SSAO = Ultra, Shadow Quality = Ultra, Mirrors Reflection Quality = High, SSR Quality = High, Variable Rate Shading = Quality.
• Horizon Zero Dawn. 1920 × 1080 resolution: Preset = Ultimate Quality. 3840×2160 resolution: Preset = Ultimate Quality.
• Metro Exodus. 1920 × 1080 resolution: DirectX 12, Quality = Ultra, Texture Filtering = AF 16X, Motion Blur = Normal, Tesselation = Full, Advanced PhysX = Off, Hairworks = Off, Ray Trace = Off, DLSS = Off. 3840×2160 resolution: DirectX 12, Quality = Ultra, Texture Filtering = AF 16X, Motion Blur = Normal, Tesselation = Full, Advanced PhysX = Off, Hairworks = Off, Ray Trace = Off, DLSS = Off.
• Shadow of the Tomb Raider. 1920 × 1080 resolution: DirectX12, Preset = Highest, Anti-Aliasing = TAA. 3840×2160 resolution: DirectX12, Preset = Highest, Anti-Aliasing = Off.
• A Total War Saga: Troy. 1920 × 1080 resolution: DirectX 12, Quality = Ultra, Unit Size = Extreme. 3840×2160 resolution: DirectX 12, Quality = Ultra, Unit Size = Extreme.
• Watch Dogs Legion. 1920 × 1080 resolution: DirectX 12, Quality = Ultra, RTX = Off, DLSS = Off. 3840×2160 resolution: DirectX 12, DirectX 12, Quality = Ultra, RTX = Off, DLSS = Off.
• World War Z. Resolution 1920 × 1080: DirectX11, Visual Quality Preset = Ultra. 3840×2160 resolution: DirectX11, Visual Quality Preset = Ultra.

In all gaming tests, the results are the average number of frames per second, as well as the 0,01-quantile (first percentile) for the FPS values. The use of 0,01-quantile instead of the minimum FPS indicators is due to the desire to clear the results from random performance spikes that were provoked by reasons not directly related to the operation of the main platform components.

⇡ # Performance in complex tests

The PCMark 10 simulation of typical consumer scenarios ranks the Core i7-11700K above the Comet Lake generation, but below the Ryzen 7 5800X. The conclusion suggests itself that the Core i7-11700K lags behind the Core i9-11900K by a significant amount, since the flagship Rocket Lake surpassed in speed not only its predecessors, but also competitors from AMD. But in reality, the gap is only from 3 to 5%, which does not seem to be such a big value if you remember that the clock speed at full load, the Core i9-11900K outperforms the Core i7-11700K by 10%.

In our Core i9-11900K review, we said that 3DMark Time Spy Extreme considers the flagship Rocket Lake to be the fastest eight-core processor at the moment. The Core i7-11700K in this test is also evaluated positively, although in terms of the overall performance index it is still slightly outperformed by the competing Ryzen 7 5800X. But on the other hand, in the processor benchmark, the Core i7-11700K manages to recoup and take a higher position compared to the current eight-core AMD processor.

⇡ # Application performance

On average, in resource-intensive applications, the Core i7-11700K is 6,5% slower than the flagship processor in the new Rocket Lake series. At first glance, this does not seem to be such a serious lag, but in reality, such a gap is enough for the processor in question to take a slightly different place in the existing hierarchy. First, the Core i7-11700K is clearly inferior to the previous generation flagship, the Core i9-10900K. Secondly, it is inferior to AMD’s current eight-core processor, the Ryzen 7 5800X processor. The gap in both cases is not that large, only within a few percent, but as a result, the Core i7-11700K lags behind both the Core i9-10900K and the Ryzen 7 5800X in 11 out of 13 applications in which we conduct tests.

However, this, of course, is not a disaster, because the Core i7-11700K is cheaper than both of these processors. You just need to keep in mind that, firstly, the Core i7-11700K is slower than offerings with a large number of computing cores, both current and past generations. Secondly, the Zen 3 microarchitecture, even after the appearance of Cypress Cove, remains the leader in terms of specific performance. Therefore, when it comes to chips with a similar clock speed, as in the case of the Core i7-11700K and Ryzen 7 5800X, the AMD option will be faster, at least when it comes to work in resource-intensive applications for creating and processing digital content.

Rendering:

Photo processing:

Work with video:

Video transcoding:

Compilation:

Archiving:

Chess:

Encryption:

⇡ # Performance in games. 1080p tests

Based on the results of gaming tests, it becomes clear why the Core i9-11900K and Core i7-11700K, which are similar in many basic characteristics, were scattered across different series. The point is simply that the Core i9 is sent by Intel on a mission called “the fastest processor for gaming.” The Core i7-11700K does not face any such task, its clock frequency is kept within reasonable limits, and as a result, it is inferior in average FPS to its older brother by about 3,5%.

All of this can be clearly seen in the following graph, which summarizes the performance test results in a dozen popular gaming applications.

In general, most modern AMD and Intel processors have fairly close gaming performance. Serious claims in this regard can only be made against representatives of the Ryzen 3000 series.However, to be precise to the end, I have to say that the gaming performance of the Core i7-11700K is slightly better than that of the Comet Lake generation processors, but worse than that of Ryzen 9 5900X processor. As for the eight-core Ryzen 7 5800X processor, it turns out to be almost identical in average gaming performance with the Core i7-11700K.

⇡ # Performance in games. 2160p tests

Increasing the resolution leads to a higher load on the video subsystem, so the effect of processors on the frame rate in 4K is not so pronounced. And in this case, between Comet Lake, Rocket Lake, and at the same time Zen 3, you can actually put an equal sign, at least until new graphics cards with higher performance than the GeForce RTX 3090 come to the market.

⇡ # Power consumption

Formally, the Core i7-11700K claims the same thermal and power characteristics as the flagship Core i9-11900K. However, its lower clock speeds give hope that it is still not as voracious as its older brother. For verification, we carried out a series of measurements of the total consumption of the test platform (without a monitor) under various load conditions. The results are presented in the diagrams.

Here you can see both positive and negative points. On the one hand, the Core i7-11700K is a significantly more economical eight-core processor against the background of the Core i9-11900K. The system with it consumes 60-80 W less under load on all cores compared to a similar platform with the installed flagship. However, such a reduction in consumption does not lead to a fundamental change in the situation: the Core i7-11700K is still much more power hungry than the Core i7-10700K and even more so than the Ryzen 7 5800X. However, the eight-core Core i7-11700K still falls short of the consumption level of the deca-core Comet Lake.

⇡ # Conclusions

Without going into details, about the Core i7-11700K we can say: “The same as the Core i9-11900K, only a little slower and noticeably cheaper.” And in general, this characteristic contains almost everything that at the consumer level is enough to know about the processor considered today. Technically, it differs from the flagship in the Rocket Lake family only in lower clock frequencies and a poorer set of turbo modes, otherwise there are no important differences. This is the same full-fledged eight-core processor, like the Core i9-11900K, it can be overclocked in the same way, and it has no restrictions in terms of the memory controller (which one might think about, based on Intel’s statements).

The 25% discount that is included in the price of the Core i7-11700K relative to the Core i9-11900K is obviously due solely to the lack of status and premium on the Core i7-11700K, because the difference in performance that exists between these processors is in reality insignificant. As tests have shown, the Core i7-11700K loses to its older brother by only 4-6% in applications and games. Moreover, given the overclocking capabilities and the available frequency potential, enthusiasts will be able to minimize this gap without much difficulty. In other words, if the choice of the most advantageous eight-core in the Rocket Lake family is approached from the standpoint of rationality, then, of course, the Core i7-11700K will win – in terms of performance and price it looks the most attractive.

However, the other hope that we pinned on the Core i7-11700K did not come true. It seemed that a more restrained frequency formula would make it not so gluttonous and hot and would make it possible to compare this processor in terms of heating at least with eight-core representatives of the Comet Lake generation. But no, even though the Core i7-11700K does not heat up as much as the Core i9-11900K, it leaves all the other eight-core processors far behind in terms of heat dissipation. In other words, this is also a very hot chip that needs efficient, and preferably liquid, cooling systems.

Nevertheless, in the end, the Core i7-11700K can still (with certain reservations) be called an acceptable alternative to AMD’s current eight-core processor, the Ryzen 7 5800X. Due to the advantages brought by the Cypress Cove microarchitecture, Rocket Lake processors have become noticeably faster than their predecessors of the Comet Lake generation with the same number of processing cores. As a result, the gaming performance of the Core i7-11700K and the Ryzen 7 5800X are comparable, while the application performance advantage of AMD is only a few percent. But Intel is ready to generously compensate for this small lag along with high heat dissipation with a price. According to official data, the Core i7-11700K is $ 7 cheaper than the Ryzen 5800 50X, plus you shouldn’t forget about the existence of an even more affordable version of the considered processor, Core i7-11700KF, in which the graphics core is deactivated.

Thus, in the segment of eight-core processors, the emergence of Rocket Lake gives Intel some chance to contain the onslaught of a competitor. Of course, we are not talking about a change of leader today, but the new processors are probably capable of slowing down the recent migration of users from the Intel platform to the AMD platform.

3DNews editors would like to thank iRU for providing the Core i7-11700K processor.