Intel’s 12th Gen desktop processors promise a giant leap, thanks to a long-awaited shift to a modern 10nm process known as Intel 7, and a new hybrid that echoes Apple’s powerful M1 chip design. It’s the most exciting Intel announcement in years, and it comes just in time: just as AMD’s Ryzen 5000 processors have taken Team Blue’s gaming crown (and significant market share).
We’ve been testing the $589 Core i9 12900K and $289 Core i5 12600K against their 11th-gen predecessors and AMD’s Ryzen 5000 processors for the past week, and have been impressed with what Intel has accomplished here. However, all that horsepower comes at a price — we’re not sure yet if the 12th generation makes sense for everyone. In this review, we’ll explain what gaming performance you can expect, explore the new features of the Z690 platform, and everything else you need to know.
Before we get into the results, let’s take a brief look at Intel’s new Alder Lake architecture. The new CPUs feature high-speed “performance” and low-power “efficient” cores, as we determined in Intel’s 12th Gen Announcement article. The idea is that Intel’s thread controller and the Windows 11 operating system work together to put performance-critical gaming or content creation tasks on the P cores, while background tasks like updating or streaming run on the E cores.
The flagship 12900K is equipped with 8 P cores and 8 E cores, and each P core also provides hyperthreading, for a total of 16 cores and 24 threads. These processors also feature a larger L3 cache, reflecting AMD’s move to significantly boost gaming performance with its Ryzen 5000 processors. Combining all the factors – the move to 10nm, new architecture, new cache and max turbo boost up to 5.2GHz – Intel promises up to 20% better gaming performance and up to 30% better content creation – well above the usual single-digit performance we’ve come to expect promote.
Intel’s 12th Gen CPUs are also appearing with the new Z690 motherboards. These support Alder Lake’s new higher CPUs with LGA 1700 sockets, as well as PCIe 5.0 support. Models supporting DDR4 and DDR5 RAM are available as they become available. We think DDR4 will prove to be the more valuable option here and now, with DDR5 motherboards coming in a few months or years. After all, the first DDR4 RAM kits were slow and high latency, but that quickly improved within the first few years of retail availability. We’ll cover RAM testing later. Now, before we jump into gaming benchmarks, let’s take a quick look at the testbed and content creation performance we’ll be using.
| processor | Color(P/E) | thread | P max turbo | E max turbo | Smart Cache | cost |
|---|---|---|---|---|---|---|
| i9-12900K | 16 (8E/8P) | twenty four | 5.2GHz* | 3.9GHz | 30MB | $589 |
| i9-12900KF | 16 (8E/8P) | twenty four | 5.2GHz* | 3.9GHz | 30MB | $564 |
| i7-12700K | 12 (8P/4E) | 20 | 5.0GHz* | 3.8GHz | 25MB | $409 |
| i7-12700KF | 12 (8P/4E) | 20 | 5.0GHz* | 3.8GHz | 25MB | $384 |
| i5-12600K | 10 (6P/4E) | 16 | 4.9GHz | 3.6GHz | 20MB | $289 |
| i5-12600KF | 10 (6P/4E) | 16 | 4.9GHz | 3.6GHz | 20MB | $264 |
Our initial tests used a DDR5 motherboard: the Asus ROG Maximus Z690 Hero. (We also planned to test DDR4 for more comparisons, but our motherboard didn’t arrive in time for the embargo. We’ll revisit this topic in the future.) We tested two dual-channel RAM kits: 32GB Corsair DDR5-5200 CL38 Dominator The Platinum kit, with fancy RGB lighting and a 32GB Crucial DDR5-4800 CL40 kit, looks more basic but costs less. Our other motherboards use the fastest DDR4 we have on hand: the 3600MHz CL16 G.Skill Trident Z Royal kit. This memory is much lower clocked, but it does offer significantly better timings and cost less too, so it will be interesting to see how it performs on the Z690 DDR4 board once we test it internally.
The Asus motherboard is one of the most expensive Z690 options out there, retailing for around £520/$600, but includes one of the thickest VRM coolers I’ve seen, an impressive 20+1 power stage and two Thunderbolt 4 ports. There are also some unique features including a quick release button for the main PCIe slot for faster removal (great for GPU testing!), screwless M.2 SSD mounting and “ROG Hyper M.2” PCIe 5.0 card – neat .
Elsewhere, we tested 11th-gen processors with the Asus ROG Maximus Z590 Hero and AMD Ryzen 5000 CPUs with the Asus ROG Crosshair 8 Hero. These CPUs were cooled with an Eisbaer Aurora 240mm AiO, while the 12th-gen tests were conducted with the included Asus ROG Ryujin 2 360mm AiO. (And to answer an obvious question: Based on our testing, 240mm and 360mm all-in-ones tend to offer the same performance – especially on open-air test rigs in cool (21C) ambient conditions. The only difference is often the fan speed, at 240mm vs 360mm.) Our rig was done using a 1000W Corsair RM1000x power supply from Infinite Computing.
To get the 12900K and 12600K to keep up with them, we upgraded from the RTX 2080 Ti we used in our 11th Gen review to one of the fastest consumer graphics cards: the RTX 3090. The RX 6900 XT was meant to be a serviceable replacement, but we have three RT-based games in our lineup, and Nvidia’s better performance in those games makes it the obvious choice. Our particular model is the Asus ROG Strix 3090 OC Edition, which features a large three-slot, three-fan design that keeps the graphics card surprisingly cool and quiet, especially compared to the more modestly sized 2080 Ti Founders Edition. This should help reduce the variance between runs and ensure we are as CPU bound as possible.
We tested on a (£196/$160) 1TB PNY CS3140 PCIe 4.0 SSD with a fresh install of Windows 11. All the latest BIOS updates and Windows updates installed, as well as AMD’s latest chipset drivers for testing Ryzen 5000 processors; again, all games used are up to date. (Note that this new platform means our results here aren’t comparable to our past Ryzen 5000 or Intel 11th Gen reviews.)
It is important to note that we tested with Multi-Core Enhancement (MCE) enabled on all Intel motherboards. With the removal of turbo restrictions in Intel’s 12th Gen guidelines, it’s clear that the chipmaker is aware that its motherboard partners — and its users — tend to use their processors that way anyway. However, this does mean that these CPUs are likely to generate more heat and consume more power than the listed power figures, so you’ll need to factor in the cost of a reasonable motherboard, a good power supply, and a robust cooling solution Obtain similar results.
Above – my favorite feature of the Z690 Hero – true innovation! Animation source: ASUS.
So, proviso – content creation. We’ve picked two relatively quick tests here to give you an idea of how Gen 11 performs common tasks like 3D rendering and transcoding video from one format to another. For the former, we used Cinebench R20, a benchmark application that simulates rendering 3D scenes in the professional graphics package Cinema 4D, and for the latter, we used high-quality Patreon video encoding to h.264 and h.264. 265 (HEVC) format free Handbrake app – a task we perform on a regular basis.
Cinebench tests single- and multi-threaded performance, which can make it a useful predictor of future gaming performance – for example, a CPU with a high single-core score here is likely to perform well in games that rely on single-threading for most of their computations, such as Far Cry 6. Intel promises massive single-core performance gains, and it’s confirmed—two of our internal CPUs scored 700 points or more in a single core, roughly 200 points higher than their predecessors and even outperformed AMD’s Ryzen 5000 designs.
Multithreading scores also benefit from the 12th Gen’s single-core speed advantage, with the 12900K surpassing the 10,000-point threshold for the first time in our testing. The 12600K even surpasses the 11900K under Windows 11, which is a very tricky situation for anyone investing in a high-end 11th-gen system at launch. However, we should keep in mind that we’re using DDR5 memory here, so the DDR4 test may show a closer apple-to-apple gap than we’ve seen here.
| CB R20 1T | CB R20 metric tons | HB h.264 | HB HEVC | HEVC power usage | |
|---|---|---|---|---|---|
| Core i9 12900K | 760 | 10416 | 70.82fps | 29.26fps | 373W |
| Core i5 12600K | 716 | 6598 | 44.27fps | 19.99 frames per second | 223W |
| Core i9 11900K | 588 | 5902 | 41.01fps | 18.46fps | 321W |
| Core i5 11600K | 541 | 4086 | 29.00 frames/sec | 13.12fps | 250W |
| Ryzen 9 5950X | 637 | 10165 | 70.28fps | 30.14fps | 237W |
| Ryzen 9 5600X | 601 | 4502 | 31.75fps | 14.43fps | 160W |
On the Handbrake bench, the 12th gen and DDR5 also look more competitive – the 16-core 24-thread 12900K is able to outperform the 16-core 32-thread in H.264 encoding, and falls only slightly behind AVX-heavy HEVC encoding . Even the 12600K is capable of nearly 10% more power than the Core i9 11900K from earlier this year, while consuming nearly 100W less power during HEVC encoding. The 12900K is still very power hungry in its “unleashed”, MCE-enabled mode, especially compared to AMD’s ultra-efficient Ryzen 5000 series, but if the performance is this good, I don’t think many people will mind.
That’s certainly impressive stuff, but now let’s take a look at the real meat and potatoes – game results. We’ve tested the new chips, their predecessors, and AMD’s closest competitors in eight games here, including several new additions to our CPU test suite. We also took a quick look at memory performance here, but we plan to do a full analysis once we get our hands on the DDR4 board. Choose your adventure from the options below, or just click the “Next” button to move on.
Intel Core i9 12900K and Core i5 12600K Analysis
