Intel Core i9-12900KS Review: Intel’s Fastest Gaming CPU Yet

If Intel’s recent acquisition of the RPS Guide to Best Gaming CPUs was not an indication enough, their latest 12th generation Alder Lake chips are the real deal – especially on pure gaming performance. This should in a way make the newly launched Core i9-12900KS the real deal of all: it is structurally identical to the Core i9-12900K, originally the top-specified CPU in the series, but made with an even more discerning binning process. A component of the highest quality components, if you will, and as such can brush past the Core i9-12900Ks maximum boost clock speed of 5.2 GHz to reach a burning speed of 5.5 GHz.

Higher speeds per core are usually good news for gaming, and it’s not just the upper end where the Core i9-12900KS makes improvements: its eight larger, poorer performance cores (P-cores) get a 200MHz base clock increase of up to 3.4 GHz, and the eight smaller Efficiency cores (E-cores) increase 100MHz up to 2.5GHz. The final results often support Intel’s claims that this is “the world’s fastest desktop processor”, although it is also a sort of CPU version of the RTX 3090 Ti graphics card: technically better, but perhaps not by a large enough margin to justify over cheaper alternatives.

Unfortunately, due to sampling delays, I was unable to test the Core i9-12900KS with the AMD Ryzen 7 5800X3D, which the red team also claims to be the planet’s fastest gaming CPU (albeit using a completely different method: upload to cache instead to maximize core speeds). What I can tell you is that the Core i9-12900KS smokes all other known Ryzen and, in fact, Intel Core CPUs as a desktop multitasker. In the Cinebench R20 reference, its single-core score of 811 and multi-core 11,019 put it far ahead of both the Core i9-12900K (756 and 10,519 respectively) and the AMD Ryzen 9 5950X (639 and 10189).

They also nicely demonstrate both the major benefits of the newer Intel chip: the higher core clock speeds are responsible for the monster’s single-core performance, while its multi-core capabilities come from the distinctive Alder Lake combination of P-cores and E-cores. As with most of its 12th generation stable mates, the Core i9-12900KS can more efficiently shunt small pieces of workload to the cores that are best suited for them – demanding work goes to the P cores, lighter jobs go to the E cores. If you also need the game PC to juggle editing or heavy coding, the benefits are obvious.

A bar graph showing the Intel Core i9-12900KS game reference results, along with the results of competing CPUs.

A bar graph showing the Intel Core i9-12900KS game reference results, along with the results of competing CPUs.

When it comes to gaming specifically, the Core i9-12900KS comes out on top more often than not. Take a look at these reference results, recorded with an RTX 2080 Ti GPU: you can see for yourself that it consistently goes ahead of both the Core i9-12900K and Ryzen 9 5950X. The latter was tested with DDR4 RAM, not DDR5 like all four of the 12th generation Intel CPUs, but I have found that DDR5 is neither significantly better nor worse than DDR4 on current hardware and games.

Aside from an incredibly good Assassin’s Creed Valhalla performance of the Intel Core i5-12400F (I’ve tested this several times, and yes, it’s legit), either the Core i9-12900KS either leads or matches its rivals in every single game. It’s definitely the fastest gaming CPU Intel has ever made; the fastest in the world? It’s up to Ryzen 7 5800X3D, but I would say that Core has a very good shot.

There is a difference, mind, between being the fastest and being the best. As with the RTX 3090 Ti, the underlying fun killer is money: when the Core i9-12900KS costs £ 720 / $ 750, about £ 120 / $ 150 more than the Core i9-12900K, you’d expect the game performance gap to reflect that difference .

An Intel Core i9-12900KS CPU is held between a finger and a thumb.

It does not. It’s well on average 10fps more than the S-loose model in Shadow of the Tomb Raider, but at 160fps and above you can barely see the difference. And in almost every other game there is either an insignificant difference of 2-3 fps, or none at all. Assassin’s Creed Valhalla extends the lead to the Core i9-12900KS a bit, but an extra 6fps is just not worth another £ 120.

And it’s just comparing it to other Core i9s. Yes, technically they tend to be a few more FPS behind, but the relatively cheap Core i5-12600K and Core i5-12400F are competitive with the Core i9-12900KS in anything but Forza Horizon 4. And maybe Shadow of the Tomb Raider, although the high frame rates in both cases mean that 10-20 fps deviations are not as visible as they would be at lower levels. Elsewhere, there are even a bunch of single-digit differences, which make the Core i9-12900KS value proposition even worse.

In addition to paying more than you need, choosing the Core i9-12900KS will also mean using more power than you need. Base power consumption – not even maximum consumption – is 150W, 25W more than Core i9-12900K and Core i5-12600K and a corrosive 85W more than Core i5-12400F. The Core i9-12900KS is unlocked for overclocking, unlike the Core i5-12400F, but it will also put you in a difficult situation with heat.

An Intel Core i9-12900KS CPU installed on a motherboard.

I tested the Core i9-12900KS with an Asus ROG Ryujin II 360, a rather expensive (but efficient) AIO water cooler. When you play, it does not get dangerously hot, with the warmer P-cores most often staying in the range 44-60 ° c. But it is up to 10 ° c warmer than the Core i9-12900K, and the P-core peak at 71 ° c was also higher. So the cheaper chip will be easier to overclock on liquid cooling, although both CPUs can also easily reach 100 ° C (what I like to call the uh-oh temperature) during Cinebench R20 runs. In stock speeds!

I would say that this is not too big an issue if you are just building a gaming PC, instead of a multipurpose workstation, but if that is the case, it’s better to forget the Core i9s and just get a Core i5-12600K (or Core i5 -12400F) instead. The Core i5-12900KS ‘lure as a carefully crafted record setter is not as strong as the satisfaction you get from saving hundreds on a processor with essentially identical performance.