A little over a year has passed since
Comet Lake-S
The code name for the 10th generation of Intel Core processors is Comet Lake. And yes, it's still 14 nm. Another refresh
socket change
Yes, nothing lasts forever under the moon, and simultaneously with the release of the 10th generation, Intel rolled out a new socket - LGA 1200 (Socket H5). Despite the fact that it is compatible in mounting holes (75 mm) with existing cooling systems, the illusory hope that they would not have to be changed disappeared after the first preliminary tests. But more on that later.
More cores, higher frequency
Already traditional for Intel way out of the situation with nanometers: if you do not change
The Intel i9-10900K processor was thrown two cores, respectively, 4 threads per
Since the process has not changed, the heat dissipation requirements, or
If the base frequency of the new processor has risen by only 100 MHz - from 3,6 to 3,7, then with
In addition to the already familiar Turbo Boost technology, a mega-super turbo boost has appeared. Officially it is called Thermal Velocity Boost (TVB). It should be noted that this technology was introduced back in the eighth generation of Intel Core, but only selected representatives got it. For example, I personally know i9-9980HK and i9-9880H.
The essence of the technology lies in the fact that at a certain temperature of the processor, the frequency of one or more cores rises above Turbo Boost. The value of the added frequency depends on how much the operating temperature of the processor is below the maximum. The maximum frequency of processor cores with Intel Thermal Velocity Boost Technology enabled is achieved at an operating temperature of no higher than 50Β°C. As a result, in TVB mode, the clock speed of one core rises to 5,3 GHz, and the remaining cores to 4,9 GHz.
Since in the new generation of cores there are two more cores, in a state of maximum auto-overclocking with all types of βboostsβ, this βstoveβ emits up to 250 W, and this is already a challenge even for a water cooling system (CBO), especially in a compact case, without removal waterblockβ¦
They talked about the cores, explained about the frequencies, complained about the socket, let's move on. The main changes include a slightly increased L3 cache and an increased frequency of supported RAM - from DDR-2666 to DDR4-2933. In general, that's all. Intel did not even update the integrated graphics core. The amount of RAM has not changed either, the same 128 GB was inherited from the previous generation. That is, as always with refreshes: they threw in cores and frequencies, however, they also changed the socket. There are no other significant changes, at least in terms of servers, no. I propose to move on to testing and see how the performance of the new generation has changed in comparison with the past.
The test is
Two processors of the Intel Core line are involved in testing:
- Ninth generation i9-9900K
- Tenth generation i9-10900k
Tactical and technical characteristics of platforms
Processors Intel i9-9900K
- Motherboard: Asus PRIME Q370M-C
- RAM: 16GB DDR4-2666 MT/s Kingston (2pcs)
- SSD: 240 GB Patriot Burst (2 in RAID 1 - a habit developed over the years).
Processors Intel i9-10900K
- Motherboard: ASUS Pro WS W480-ACE
- RAM: 16GB DDR4-2933 MT/s Kingston (2pcs)
- SSD: 240 GB Patriot Burst 2 in RAID 1.
Both configurations use single-unit water-cooled platforms. But there is a caveat⦠In order not to lose the TVB frequencies and start the Intel i9-10900K normally, I had to assemble a powerful custom water cooling system (hereinafter referred to as CBO) for the platform with the tenth generation Core. This required some effort (and a lot of it), but this solution allowed us to get a stable 4,9 GHz to each core at the most peak loads without crossing the temperature threshold of 68 degrees. Bow to the heroes-customizers.
Here I will allow myself some digression from the topic and explain that such an approach to business is dictated solely by pragmatic considerations. We find technical solutions that give maximum performance with minimum rack disposal, while getting an adequate cost. At the same time, we do not overclock the hardware and use only the functionality that was laid down by the hardware developers. For example, regular overclocking profiles, if the platform has any at all. No manual setting of timings, frequencies, voltages. That allows us to avoid all sorts of surprises. As, in fact, and preliminary testing, which we conduct before we give ready-made solutions into the hands of customers.
It is no coincidence that we always test in single-unit configurations - such testing is quite enough to make sure that the solution found is reliable. As a result, the client receives proven equipment and maximum speed at the lowest price.
Returning to our i9-10900K, I note that the temperature of none of the compared processors did not rise above 68 degrees. And this means that the solution, with other advantages, also has a good overclocking potential.
Software part: OS CentOS Linux 7 x86_64 (7.8.2003).
Kernel: UEK R5 4.14.35-1902.303.4.1.el7uek.x86_64
Made optimizations relative to the standard installation: added kernel launch options elevator=noop selinux=0
Testing was done with all Specter, Meltdown and Foreshadow patches backported to this kernel.
Tests used
1. Sysbench
2.geekbench
3. Phoronix Test Suite
Detailed description of tests
Geekbench test
A suite of tests conducted in single-threaded and multi-threaded mode. The result is a performance index for both modes. In this test, we will look at two main indicators:
- Single-Core Score - single-threaded tests.
- Multi-Core Score - multi-threaded tests.
Units of measurement: abstract "parrots". The more "parrots", the better.
Sysbench test
Sysbench is a package of tests (or benchmarks) for evaluating the performance of various computer subsystems: processor, RAM, data storage devices. The test is multi-threaded, for all cores. In this test, I measured one indicator: CPU speed events per second - the number of operations performed by the processor per second. The higher the value, the more efficient the system.
Phoronix Test Suite
Phoronix Test Suite is a very rich test suite. Almost all the tests presented here are multi-threaded. The only exceptions are two of them: single-threaded Himeno and LAME MP3 Encoding tests.
In these tests, the higher the score, the better.
- John the Ripper multi-threaded password guessing test. Let's take Blowfish cryptalgorithm. Measures the number of operations per second.
- The Himeno test is a linear Poisson pressure solver using the Jacobi point method.
- 7-Zip Compression - A 7-Zip test using p7zip with an integrated benchmark feature.
- OpenSSL is a set of tools that implement the SSL (Secure Sockets Layer) and TLS (Transport Layer Security) protocols. Measures the performance of RSA 4096-bit OpenSSL.
- Apache Benchmark - The test measures how many requests per second a given system can handle when executing 1 requests, with 000 requests running concurrently.
And in these, if less, then better - in all tests, the time it takes to pass is measured.
- C-Ray tests CPU performance on floating point calculations. This test is multi-threaded (16 threads per core), will shoot 8 rays from each pixel for anti-aliasing, and generate a 1600x1200 image. The test execution time is measured.
- Parallel BZIP2 Compression - The test measures the time required to compress a file (Linux kernel source .tar package) using BZIP2 compression.
- Audio data encoding. The LAME MP3 Encoding test is performed in one stream. The test time is measured.
- Video encoding. The ffmpeg x264 test is multi-threaded. The test time is measured.
Test results
i9-10900K is better than its predecessor 44%. In my opinion, the result is simply gorgeous.
Difference in single-threaded test of everything 6,7%, which is generally expected: the difference between 5 GHz and 5,3 GHz is the same 300 MHz. It's just 6%. And there were some conversations π
But on the other hand, in the multi-threaded βparrotsβ test, the novelty has almost 33% more. Here TVB played an important role, which we were able to use almost to the maximum with a custom CBO. At the peak, the temperature in the test did not rise above 62 degrees, and the cores worked at a frequency of 4,9 GHz.
Difference 52,5%. Just like in the Sysbench and multi-threaded Geekbench tests, such a significant lead is achieved due to CBO and TVB. The temperature of the hottest core is 66 degrees.
In this test, the difference between processors of different generations is 35,7%. And this is the same test that keeps the processor under maximum load 100% of the time, warming it up to 67-68 degrees.
97,8%. The probability of almost twofold superiority due to 2 cores and a few megahertz is βextremely smallβ. Therefore, the result is more like an anomaly. I assume that either the optimization of the test itself or the optimization of the processor takes place here. Or maybe both. We will not rely on the results of this test in this case. Although the figure is impressive.
And here I am absolutely sure that optimization was made in the test itself. This is also proven by repeated tests of AMD Ryzen, which pass it much better, despite the fact that Ryazans are not so strong in single-threaded tests. Therefore, the advantage in 65% will not count. But it was simply impossible not to talk about it. Nevertheless, we write one - we keep two in mind.
The difference between generations 44,7%. Everything is fair here, so we count the result. After all, this is exactly the test in which the maximum performance is squeezed out in a single-threaded load. On the one hand, here you can see the work done on finalizing and optimizing the kernel - a refresh with a refresh, but something under the hood was clearly optimized. On the other hand, such results may indicate that we failed to squeeze the maximum last time in the same test with the i9-9900K. I would love to read your thoughts on this in the comments.
The tenth generation confidently overtakes the ninth by 50,9%. Which is quite to be expected. The cores and frequencies added by the Intel i9-10900K rule here.
The difference between generations 6,3%. In my opinion, the result is rather controversial. In future articles, I'm thinking of dropping this test altogether. The fact is that in systems with more than 36 cores (72 threads), the test does not pass at all with standard settings, and the difference in results sometimes has to be calculated by the third decimal place. Well, we'll see. You can share your opinion on this matter in the comments.
Difference on 28%. There are no surprises, anomalies and optimizations noticed. Pure refresh and nothing more.
i9-10900K overtake i9-9900K by 38,7%. As in the case with the results of the previous test, the difference is expected and well shows the real gap between processors on the same microarchitecture.
So, let's sum up. In general, nothing unexpected - the i9-10900K outperforms its predecessor i9-9900K in all tests. Q.E.D. The price for this is heat dissipation. If you are looking for a new processor for home use and are going to squeeze the maximum performance out of the tenth generation Core, I recommend thinking about the cooling system in advance, because coolers alone will not be enough here.
Or come to us for dedikov. A ready-made solution on a good platform and with a very decent CBO, which, in addition to all other advantages, as we found out, also has overclocking potential.
Dedicated servers used in testing
Source: habr.com