[nextpage title=”Introduction: “] It feels like it was just yesterday that we presented and reviewed the i9-11900K and i5-11600K. Almost 7 months later, it’s the new generation of processors Intel Alder Lake which has arrived. In the end, Intel did not wait long to offer what we can call THE new generation, because indeed, we expect a lot from Alder Lake. Therefore, it is with great excitement that we received two Alder Lake processors: the i5-12600K and the i9-12900K.
Obviously, we aren’t able to offer you everything on release day. So we made a choice. Today we’ll have an article on Alder Lake’s i9-12900K and i5-12600K performance, and at 5pm we’ll review and test the ROG Maximus Z690 APEX motherboard. Articles on the MSI MPG Z690 Carbon Wifi and the Maximus Z690 EXTREME will be available early next week. We’re also currently preparing a full review about gaming performance.
This new generation has forced us to make choices. Between the discovery of these hybrid processors, our desire to test them under both Windows 10 and 11, as well as being able to get a first glance at their overclocking potential, some details had to be left aside for the moment. We haven’t tested them in games here, but we’ll do a full dedicated review about it in a later article.
In the next few days we will also begin testing DDR5 kits to give you a performance comparison and benchmark, such as we do with DDR4.
A very special thanks to the brands (Asus, MSI, GSKill and Corsair) who responded to the launch of Alder Lake and to Intel France for providing the two CPUs, and thus making this article possible. [nextpage title=”Alder Lake update: “]
A hybrid architecture :
Alder Lake features a new architecture that can be called “hybrid” since it will mix two types of cores. If you’re a hardware fan, you know that the i9-12900K (F) will have 16 cores and 24 threads. But beware, the 16 cores are actually divided into 8C + 8c.
There are the first 8 cores (P-core also called Golden Cove) sitting next to 8 other cores (E-core also called Gracemont). These two different types of cores have different roles. The P-cores, for Performance-cores, are meant to perform the tasks requiring high computing power in the foreground and single-threaded tasks, and the E-cores, for Efficient-cores, are optimized for handling scaling highly-threaded workloads and minimizing interruptions for lighter tasks and automation tasks in the background. Now both types of cores will work together depending on the task.
P-cores are 8 distinct physical cores, while E-cores are grouped in clusters of 4. We will see later that if we want to change the frequency of the E-core, it’s actually on all 4 cores at the same time that the change is made.
The frequency of the P-core and the E-core is of course different. In the case of our i9-12900K (F), the P-core frequency is 4900 MHz (with a boost clock of 5200 MHz) and the E-core frequency is 3700 MHz (with a boost clock of 3900 MHz). It’ll also be interesting to see how these different cores will behave in overclocking.
Finally, on the graphics side, we will find Intel’s Xe architecture with 32 EU for desktop processors and 96 EU for mobile and ultra mobile. Regarding power consumption, we will have processors with a TDP ranging from 9 to 125 watts. Alder Lake will also introduce new I/O technologies such as PCIe 5.0 with 16 lanes (and 4 lanes in 4.0), Wi-Fi6E and Thunderbolt 4.
A splendid media box :
A quick word on Intel’s kit packaging offered to the media for the launch of their new products. The launch of a new platform and processors often means review samples for the media. So, in addition to the motherboards and DDR5 kits, we received Intel’s media kit. We received a box with dimensions of 270 x 265 x 65 mm, stamped with Intel’s logo. Once the cardboard protection is removed, we finally discover the box itself with Intel’s latest version of its logo. Let’s see what’s hiding inside this box.
The first thing that jumps out is a cardboard printing of Intel’s i9-12900K die, with its 8 performance cores and its 8 efficiency cores. It’s a really nice reproduction which will be perfect in our lab as a new decoration item.
Inside, we can read the following message engraved on the lid : « Built for the next generation of gaming ». This first part of the box is gorgeous, and still, no sign of the CPUs so far. Underneath, Intel had thought of using a support made of plastic, in order to hold the die vertically. This support is inserted in a foam block. When we remove this block, it’s not one but two precious CPUs which are revealed !
Yes, Intel has indeed sent two processors ! We were provided with their i9-12900K and their i5-12600K, ready to be tested. We are looking forward to installing them on our Z690 chipset configurations.
The Alder Lake processor line:
We have received two Alder Lake processors: the i9-12900K and the i5-12600K. Nonetheless, Intel’s current line is broader, in this case six SKUs. The i9-12900K (F) with 16 cores and 24 threads, the i7-12700K (F) with 12 cores and 20 threads and finally the i5-12600K (F) with 10 cores and 16 threads. The six models will have 30, 25 and 20 MB of L3 cache respectively, and a TDP of 125W (and PL2 of 241W), which is slightly lower than previous generations, that could go up as high as 250W.
As a reminder, the “K” versions have an unlocked multiplier. It is therefore very easy to overclock them. The “KF” versions, on the other hand, have the integrated graphics part deactivated.
A CPU with an updated design :
Alder Lake benefits from the long awaited 10 nm process node. This evolution has an impact on the DIE’s thickness. As shown in the screenshot below, we can see the evolution over the last generations. The DIE’s height decreases as well as the thermal compound’s.
On the other hand, the chip features a thicker integrated heat spreader (IHS), which should improve the cooling of the DIE and provide better heat spreading. As a reminder, the i9-12900K DIE’s dimensions are 10.5 x 20.5 mm with a surface area of 215.25 mm². The IHS is soldered to the DIE with Thin STIM, i.e. solder thermal interface material.
New chipset = new socket :
It is not always the case, but this time it is, Intel’s new chipset, the Z690, will be the only one (at first) to support Alder Lake CPUs. The socket will be different, larger and will be called LGA-1700.
Available above, a screenshot detailing the difference between the two sockets (credit to Igor’s Lab). The LGA-1700 socket has 1700 pins, when the previous generation had 1200. This change suggests you may need to invest in a new cooling system. Some brands will provide adapters for your current cooling solution. Make sure you contact them and ask them about this, as it is most of the time given for free.
Intel will later offer other CPUs using LGA-1700 format as well as other less expensive chipsets (H670, B650 and H610) which might be however, DDR4 compatible only. This will save users from having to buy the new DDR5 to enjoy their new Alder Lake CPU. A word of caution though, some Z690 motherboards will also be available with DDR4 only. So be sure to double check if it’s DDR4 or DDR5 compatible, as it won’t be both at the same time.
[nextpage title=”What’s new in DDR5: “]
The Z690 chipset release coincides with the release of the new DDR5 memory. It promises greater bandwidth thanks to higher frequency, increased capacity and lower power consumption. But be sure to keep in mind that it is not possible to use DDR5 modules on a DDR4 motherboard and vice versa. As you’ll see below, the dimm slots are different and this design change is there for a reason, which is to avoid any disaster.
The arrival of the PMIC module:
The great novelty found on DDR5 memory is the presence of a PMIC (Power Manager Integrated Circuit) module. Its purpose is to regulate the power supplied to the memory stick. Therefore, we can find a PMIC on every single strip. It will supply itself through 5 volts rail of your motherboard and convert it to provide the required voltage for the kit to run. For example, if your kit needs 1.1 volts, each chip will receive 1.1 volts.
Previously, the motherboard was responsible for this. In reality this has pros and cons. The pros are that manufacturers no longer have to provide this control through the motherboard PCB. This means manufacturers can reduce the cost of motherboards.
Unfortunately, there are two types of PMIC. One is locked and will only accept the default values of your kit, this is called secure mode. The other type of PMIC can be tuned and will therefore allow the voltage to be increased as high as 2 volts, and even more. Fortunately, Asus on its Z690 APEX, has already started to bypass the secure mode restriction and managed to reach 1.435 volts. We cover this topic more thoroughly in our article about the APEX.
The other question is the heat released by this chip, and the additional cost for a DDR5 kit. As far as we are concerned, we did not notice any additional heat release during our tests. Regarding pricing, it will definitively be more expensive to buy DDR5, and we hope DDR5 remains widely available.
Higher frequencies and lower voltage :
So what’s changing in regards to frequency and power ? Firstly, manufacturers seem to favour 16 GB arrays and therefore dual channel kits in 2 x 16 or 2 x 32 GB when we have been used to 2 x 8 GB with DDR4. The frequencies are also on the rise with kits that start at 4800 MHz, up to 7000 MHz with G-skill. On a side note, the most common with DDR4 is 3200 MHz. Finally, high frequencies mean (much) more relaxed timings.
The voltage is improved as it can be reduced to 1.1 volts for DDR5 compared to 1.2 volts for DDR4. We are talking about a voltage of 1.1 volts for DDR5 kits at 4800 MHz. The voltage will be higher if the frequency increases as you will see below.
ECC (Error Correcting Code) compatible with DDR5
DDR5 arrays will be able to integrate ECC (Error Correcting Code). Uh, what did you say ? The idea of this module is to detect possible errors and analyse them before sending them to the CPU. Now, this module won’t be available on all DDR5 memory as this will increase the price. While we’re on the subject of price, they will unfortunately be (much) higher.
XMP 2.0 (DDR4) and XMP 3.0 (DDR5) changes:
With DDR5, we are moving from XMP 2.0 which are DDR4 profiles, to XMP 3.0. What this mean is the possibility for a manufacturer to provide three different XMP profiles for its memory kit. These profiles must be enabled in the motherboard’s BIOS.
The other big development is the ability to setup your own XMP. To put it simply, you can fine tune your own timings and voltages, and then write these settings in the memory to make your own XMP profile. Once you “save” the profile into your memory, you will can then choose this XMP profile in the BIOS, the same way you choose one of the manufacturers’ one. This is of course a feature that we will cover in detail in our upcoming DDR5 kit comparisons and benchmarks.
DDR5 kits in our lab :
Before we could begin our tests, we received several 32 GB kits. We will briefly present these kits we will test them individually in our memory kits comparisons. “Rendez-vous” in a few weeks to discover and learn more about them.
GSKill Trident Z5 5600 MT/s 36-36-36-76 at 1.20 volts :
The first kit which arrived in the lab is a GSKill Trident Z5 clocked at 5600 MHz with 36-36-36-76 timings, rated at 1.20 volts. Our sample is a 2×16 GB memory sticks, for a total of 32 GB. This kit is black and has no RGB. We are dealing here with full black design.
Corsair Dominator Platinum RGB 5200 MT/s 38-40-40-78 at 1.25 volt :
The second kit is a model which we have already tested in its DDR4 version. It is the Dominator Platinum RGB kit clocked at 5200 MHz with timings of 38-38-38-84, rated at 1.25 volts. Our sample is a 2×16 GB memory sticks, for a total of 32 GB. Aesthetically, it looks like the DDR4 version. This is the kit we will use for the comparative tests.
Kingston Fury 5200 MT/s 40-40-40 at 1.25 volt :
Last kit to arrive in the lab. We were able to get our hands on these one thanks to MSI. It is also a 32 GB kit, in a 2×16 GB memory sticks. The kit is clocked at 5200 MHz with 40-40-40 timings. [nextpage title=”Test configurations: “]
Yes, we had to choose one for the Z690 chipset and we chose the ROG Maximus XIV Z690 APEX, simply because it was the first motherboard we received.
The memory kit issue:
So, which memory kit should we use for our testing ? This is a question we asked you on our Twitter account and as a reminder, here’s what we came up with. The idea was to know what memory frequency your configuration is or will be
So we decided to start with a frequency of 3600 MHz with timings in 16-16-16-36. For the tests on the AMD platform, the FCLK will be 1800 MHz, keeping the 1:1 ratio synced. For the Intel Z490 and 10900K platform, we have enabled the Thermal Velocity Boost (TVB) through the BIOS and for the Z590 platform with Rocket Lake processors, ” Adaptive Boost Technology “. The idea is to take advantage of the maximum frequencies of the different cores of our processors. For the rest, apart from the loaded XMP profile, we left all other BIOS settings in auto mode.
Our operating system, Windows 10, is up to date with the latest updates. As far as the Alder Lake configuration is concerned, we chose the kit with the lowest frequency in order to keep a logic when choosing the 3600 MHz DDR4 kit. The idea is to have a frequency that could be the most representative of what you have at home, or will buy. Our choice was a 5200 MHz CL38 kit.
The other question was : Should we use Windows 10 or Windows 11 ?
We couldn’t find an agreement, so you’ll be able to see the performance under both OS !
The Intel Alder Lake configuration:
- Motherboard: Asus Maximus XIV APEX (BIOS 0702)
- Processor: Intel i9-12900K and Intel i5-12600K
- Memory : 32 GB Corsair Dominator Platinum RGB 5200 MT/s 38-40-40-78
- Graphics Card : NVIDIA RTX 3070 Founders Edition (test)
- Cooling system : Optimus Foundation + 480 mm EK radiator
- SSD : Western Digital SN850 NVMe SSD M.2 4.0
- Power supply: be quiet! Dark Power PRO 12 1500w
- Operating system: Windows 11
- Display: Asus PG27UQ
Intel Rocket Lake configuration :
- Motherboard: Asus Maximus XIII APEX (BIOS 0611)
- Processor: Intel i9-11900K and Intel i5-11600K
- Memory : 16 GB GSKill TridentZ Royal 3600 MHz 16-16-16-36
- Graphics card : NVIDIA RTX 3070 Founders Edition (test)
- Cooling system : Optimus Foundation + 480 mm EK radiator
- SSD : Western Digital SN850 NVMe SSD M.2 4.0
- Power supply: Corsair AX1600i
- Operating system: Windows 10 64-bit
- Display: Asus PG27UQ
Intel Comet Lake configuration:
- Motherboard: Asus Maximus XII APEX (BIOS 2004)
- Processor: Intel i9-10900K
- Memory : 16 GB GSKill TridentZ Royal 3600 MHz 16-16-16-36
- Graphics Card: NVIDIA RTX 3070 Founders Edition (test)
- Cooling system : Optimus Foundation + 480 mm EK radiator
- SSD : Western Digital SN750 NVMe SSD M.2 3.0
- Power supply: Corsair AX1600i
- Operating system: Windows 10 64-bit
- Display: Asus PG27UQ
AMD configuration:
- Motherboard: ROG Crosshair VIII Dark Hero
- Processor: AMD Ryzen 5 5600X, Ryzen 7 5800X, Ryzen 9 5900X and Ryzen 9 5950X
- Memory : 16 GB GSKill TridentZ Royal 3600 MHz 16-16-16-36 (FCLK 1800)
- Graphics card: NVIDIA RTX 3070 Founders Edition (test)
- Cooling system : EK AMD Ryzen Edition + 480 mm EK radiator
- SSD : Seagate FireCuda 520 NVMe SSD M.2 4.0
- Power supply: Super Flower Platinum 1600 watts
- Operating System: Windows 10 64-bit
- Display: Asus PG27UQ
[nextpage title=”The frequencies reached :”]
As with every new CPU we test, our first idea is to check if the claimed and marketed frequencies can actually be reached. To do so, we will use Cinebench R23 benchmark. We test using the two benchmark computation, first the Single-Core benchmark and second the Multi–core benchmark.
Test with our i9-12900K :
So, will the i9-12900k live up to everyone’s expectation ? Let’s see.
First step, we leave every setting in AUTO in the BIOS except for the XMP profile which we load to run the frequency of 5200 MHz 38-40-40-78 of our memory kit. We do not make any other changes in the BIOS. The idea is, as we said, to see if the frequencies are reached, out of the box.
Let’s start with a single core test in Cinebench R23. As you can see, we have several monitoring software such as Core temp or Hwmonitor to check the frequency.
No issue, we can see that two cores are regularly reaching 5.2 GHz, which is the boost frequency of our i9-12900K. It is the 4th core which reaches this frequency. In Multi-core, our i9-12900k also reaches the frequency it promises, with 4900 MHz on all of the eight performance cores.
Test with our i5-12600K :
Same goal with the i5-12600K. In single core, the 4900 MHz is reached on core 0. You can also notice that it is the one which is the hottest but this is excepted.
Finally, in multi-core, we reach a frequency of 4500 MHz on all six cores. We’ll talk about this later, but like the P-core, the E-core can also be overclocked, but only to a limited extent.
Now that we know that our two processors are fully functional, we can move on to our various other benchmarks. [nextpage title=”Cinebench/Geekbench Benchmarks:”]
Let’s start with one of the must of CPU benchmarks, the Cinebench series. We have chosen Cinebench R15, the R20 version, which is also more and more used, and finally Cinebench R23, the latest one. The scores will be given in Single-Core and Multi-Core.
Since Intel has made great announcements regarding the performance jump with its new CPUs, we will see what the results are in comparison to its previous generation, but also in comparison to its competitors.
To do so, we will use Benchmate, which at the time of our tests was in version 0.10.8. But we also tested with Benchmate beta version 0.10.9 which fixes some small bugs when using Cinebench R15. The other benchmark software we used is GeekBench. Those of you who follow us regularly will know that this is a test we are particularly fond of, and that we like to use when comparing different memory kits. Two versions will be used, Geekbench 3.4.2 and Geekbench 5.2.5, which both also have a Single-Core and Multi-Core score.
Cinebench R15 SingleThread and MultiThread :
A very popular benchmark especially by overclockers in order to compare the performance of different CPUs, but also to to measure the influence of some settings, whether it is positive or negative.. It also allows you to evaluate the optimizations of your OS, as well as your memory kit, by setting a frequency and trying to reach top scores by making as many points as possible.
Scores will be given in Single-Core and Multi-Core. Even if it is a little less used now compared to years ago, it remains a must for comparing CPUs performance, as it is one of the earliest benchmarks used by the community.
Oh… first benchmark and first sweat. We get a score of 179 points in Single-Core and 1259 in Multi-Cores with our i9-12900K. Uh, how can I put this, there’s something wrong here…
Actually, there is a bug when using Alder Lake and Windows 10 with Benchmate version 0.10.8. It’s the E-core score which is taken into account. A few days before the end of the NDA, a beta version 0.10.9 was released which solved this problem. The scores are now fixed and on point.
In Single-Core, it is our i9-12900K which reaches the top and gets the first place, ahead of the Ryzen 9 5950X. Also not far behind, notice the i5-12600K in comfortable third place.
In multi-threaded mode, the Ryzen 9 5950X still holds the top spot, but is not far from the i9-12900K.
Cinebench R20 SingleThread and MultiThread :
Cinebench R20 offers similar tests to the R15 version with Single-Core and Multi-Core benchmark computing. Released in March 2019, it offers a more complex benchmark than the R15 version which had become a little too easy for CPUs with many cores. Cinebench R20 requires eight times more computing power than the R15 version, and four times more memory. It can almost be considered as a stress benchmark as it requires your CPU to be stable because of its longer duration.
The Alder Lake advantage is confirmed in Single-core with our two CPUs taking the top two spots.
Our i9-12900K manages to take the lead over the Ryzen 9 5950X in the Multi-core benchmark by a pinch. Notice the position of the i5-12600K, which is ahead of all CPUs from the two previous generations of Intel’s.
Cinebench R23 Single-thread and Multi-thread :
How does this new version makes a difference compared to the previous ones? First of all, it is designed to be closer to real word use when it translates into scores, depending on what CPU you are using.
As a reminder, Cinebench only tests the capabilities of your CPU in single-thread or multi-thread. But the main difference with this benchmark, is a feature which makes the test lasts 10 minutes by default ! MAXON’s idea with Cinebench R23 is to provide a benchmark that will allow to test a certain stability in terms of temperature and boost frequencies. We will test there the quick version.
The different versions of Cinebench give similar results both with our i5-12900K and i9-12600K, dominating the other CPUs.
Same applies for multi-core, with the advantage going to team blue, and again, the i5 is able to perfectly position itself against the Rocket Lake and Comet Lake.
Geekbench 3.4.2 Single-Thread and Multi-Thread :
This benchmark is available in several revisions, version 5 having just been released a few weeks ago. After a series of different computing algorithms, you get two scores : one in single-thread and one in multi-thread. The version used for the tests is 3.4.2.
The trend continues and meets our expectations performance wise, the Alder Lakes have made a huge performance leap in single-thread compared to the two previous generations.
Nothing to report in the multi-thread benchmark, where the ranking looks the same as the previous ones.
GeekBench 5.2.5 Single-Thread and Multi-Thread :
This is the latest version of the GeekBench software, and is becoming more and more widespread, even being used now by tech journalists. It allows, just like Geekbench 3, to assess memory and processor performance through a sequence of computation benchmarks.
Latest and most widely used benchmark, in single-thread the i9-11900K stands out at the top between the i9-12900K at 1st place and the i5-12600K reaching 3rd place. .
In multi-thread, our i9-12900K blows everything away, taking no less than 4010 points off the Ryzen 9 5950X. Our i5-12600K manages to have its place on the podium, overtaking the Ryzen 9 5900X for the first time. [nextpage title=”Synthetic benchmarks:”]
Let’s start with a series of synthetic tests that we use regularly. We have opted for the CPU scores of the 3DMark benchmark series. The idea is to isolate only the CPU score by making using the custom test feature.
Here are the results obtained in Fire Strike and Time Spy. In Fire Strike, whether it is for the “standard” or “ultra” version of the software, the physics test, which assess CPU performance, is identical. As a result, you will only see the Fire Strike score. However, for Time Spy and Time Spy Extreme, the tests are different, so you will get both scores.
CPU score in Fire Strike :
Let’s start with 3DMark Fire Strike, which is one of the most used benchmarks today. It consists of two graphics tests, a CPU test and a fourth test that combines GPU and CPU at the same time. Don’t forget to disable the demo before launching the test, as it does not add anything in the final score, and makes the run longer for nothing other than making it longer, which has no use for us.
This first CPU score in Fire Strike sets things straight with the first place awarded to the i9-12900K ! This is the first and the only Intel processor that has been able to take the lead over the top three Ryzen CPUs.
Time Spy CPU Score :
The second test is 3DMark Time Spy. Although this one is running in 1440p, it has the particularity of using DirectX 12. It consists of two graphics tests and a CPU test. Just like with Fire Strike, don’t forget to disable the demo. These first two benchmarks are provided by UL Benchmark.
We have to admit that we have done the test several times to make sure nothing was wrong. Yes, as you can see we are well over 19K points for the i9-12900K leaving the other processors far behind, very far behind.
Time Spy Extreme CPU Score :
The second test is 3DMark Time Spy Extreme. Although this one is done in 2160p, it has the particularity of using DirectX 12. It consists of two graphics tests and a CPU test. Just like with Fire Strike, don’t forget to disable the demo.
This latest CPU score confirms once more the trend, the i9-12900K is the new king of the 3DMark benchmarks due to its almost insulting performance. We can’t wait to put it under some cold now. [nextpage title=”Benchmark AIDA64: “]
For the following test, which is going to be done with the AIDA64 software, we will compare the two platforms, AMD and Intel, since we used the same kit at the same frequency and with the same timings. AIDA has just seen its beta version evolve to 6.32.5659. It can be downloaded directly from the site but in order to benefit from all its features and have all of the desired results, you will need a licence.
AIDA 64 :
It is a software capable of providing a lot of useful information about your configuration, but it has the a special feature which is widely used, the memory benchmark. This will calculate the read, write and copy bandwidth as well as the memory latency of your memory kit. It is very often used to compare memory kits between them, or when overclocking your memory and see what results you get out of it.
It is also an excellent tool if you want to simply compare the performance of your memory kit with those we have already tested and given the results. The version used for testing is v6.50.5806 beta.
BOOM! Not only is the reading bandwidth really high with the i9-12900k, but also with the i5-12600K. Without a doubt, the gain from DDR5 is huge and allows increased reading speeds.
In writing, same boat, speeds are taking off big time. Now it will be very interesting to see where does an Alder Lake + Z690 + DDR4 combo fits in compared with our DDR5.
As well as copy, same observation. It’s when it comes to memory latency things get flipped upside down. Now, we can assume that the 5600 or 6000 MHz kits will certainly have lower latency.
[nextpage title=”Consumption: “]
Can we talk about a sensitive topic ? Recently, we announced in the news, the possibility of using the PL2 mode (max power consumption 241W) instead of the PL1 mode (TDP 125W). This has of course a direct impact on performance and therefore on scores.
On our side, we did not change anything and we wanted to see how our configuration would behave out of the box. We basically installed our CPUs and loaded the XMP profile. Then we did our series of benchmarks and compiled our findings for you in this article.
We wanted to measure power consumption and we decided to use Cinebench R23 benchmark. So we will be testing power usage in multi-threading. The idea is to measure the average power consumption of our different CPUs.
Please be aware that we are talking about the overall power consumption of the build, and not the CPU alone. Currently, we are not yet able to measure independently and correctly the consumption of the CPU. But this is in progress.
As you can see, the average power consumption of our configuration is 321.1 watts. Compared to the previous generation i9-11900K, this is much better. Let’s not forget that we are dealing with an 8C + 8c processor. Against the Ryzen 9 5950X, our Alder Lake does not manage to match its level of power efficiency, even if the difference is not that big.
The i5-12600K, however, is a the great surprise, as its overall power consumption is only 214.2 watts. Thanks to our tests, it gets our approval and we had a lot of good things things to say about it. [nextpage title=”And under Windows 11:”]
It was obvious to us not to avoid Windows 11 if we wanted to test Intel’s Alder Lake processors. Intel has made it very clear during its conferences that Windows 11 is there to make the most of this new generation of CPUs. Nevertheless, we tested our two CPUs under Windows 10. Why did we do this? Simply to remain consistent in our comparison chart. All of our other CPUs have been tested under this OS.
Nevertheless, we installed Windows 11, for the first time, in order to compare our scores with the 12900K. The idea was to see if there were any differences performance wise. In the next few days we will be setting up a new platform for GPU and memory testing, and in the future we will opt for Windows 11.
As you can see, the differences are very small and the switch to Windows 11 for these different benchmarks has almost no impact. [nextpage title=”Overclocking:”] Overclocking, our passion. It is of course on this point that we spent many hours to understand the subtleties of Alder Lake and DDR5. But we won’t reveal everything here, but rather tease it out to prepare you for the 17-hour test of the ROG Z690 APEX. All tests were performed on this card. In order to keep the suspense going until 5pm, here is a series of captures that will keep you waiting and coming back in a few hours.
i9-12900K and i5-12600K :
Regarding the i9-12900K, it will be possible to overclock it in two ways to submit your scores to Hwbot. Either by keeping all the cores or by disabling the E-core. The maximum frequency we were able to reach is 5300 MHz on the 8 Performance cores at 1.29 volts. As you can see, it heats up a lot. We can only recommend a custom watercooling solution.
The i5-12600K gave us a very pleasant surprise. We went from 4.5 GHz to 5.1 GHz (1.28 volts), while maintaining an acceptable temperature. It should be noted that we are on a custom cooling solution based on a 480 mm radiator.
GSKill 5600 MHz CL36 to 6600 MHz CL38 kit:
Overclocking the DDR5 also went very well. The presence of profiles on the Z690 APEX is a really great feature. We have quite a good number of scores to share with you, but you’ll have to wait until tomorrow for that.
A bit teasing though: we pushed one of our kit to 6600 MHz CL38, but did not find stability yet. However, we also overclocked a 6000 CL32 and this has given us huge performance.
See you at 5pm for a detailed look at the overclocking of the 12900K, 12600K and the GSKill Trident Z5 5600 CL36 kit. We will of course present and detail the BIOS options and the settings used. [nextpage title=”The last word:”]
Well, that’s it, Intel is back in the game and in a big way! Because yes, let’s be honest, we’ve been waiting for a new architecture from Intel for several years. The bet has been successful with the i9-12900K which takes the lead on all of our benchmarks and the i5-12600K that is very often ahead of the top of the range Rocket Lake and Comet Lake.
Single-core performance has increased significantly compared to the previous generation and is now ahead of the Ryzen 5000’s. Regarding power consumption, the switch to 10 nm is positive even if there is still a significant gap with the Ryzen 9 5950X. But that is different with the i5-12600K, which is the one offering the best balance between performance and power consumption. We strongly believe the i5-12600k is the new best of class the 6-core CPUs, especially thanks to its affordable price and its performance.
The i9-12900K becomes “the processor” for 3DMark. The CPU manages to pulverize records in every benchmark, making it the most interesting SKU for benchmarkers. Nevertheless, Cinebench and Geekbench are not left out either.
Even if we haven’t covered the topic of games yet, there’s no doubt that the performance will be up to scratch and will confirm your interest in these two CPUs, with the i5-12600K being more the budget friendly one.
Alder Lake, interesting prices :
Should I change my setup and go for Alder Lake? Given the performance and if you haven’t just bought yourself a brand new setup, the answer is clearly yes. When you see the i5-12600K available for 330 euros, it would be a shame not to go for it. Some concerns still remain and that’s the rest of the component’s prices.
The first Z690 motheboards are very expensive and the lack of DDR5 kits availability is also paying off. We’ll try to pick up a DDR4 compatible Z690 board soon to see if this option could be the better deal for now.
The i9-12900K costs around 720 euros. It is therefore more interesting than the Ryzen 9 5950X sold for instance for around 850 euros. We will see if AMD decides to change its pricing in the coming weeks. The good news is that Intel confirmed there will be sufficient availability for its Alder Lake processors. In conclusion, you might have guess it yet. In conclusion, Alder Lake is a big WIN and we say YES.
See you at 5pm to discover its overclocking potential on the ROG Maximus Z690 for the CPU but also for DDR5!