New Intel 13900K & 13600K - HUGE Power

The fight between Intel and AMD continues, and honestly, I can't remember a time when the results have been this close. Both companies are chasing that 6GHz clock speed barrier, both of them have adopted the DDR5 memory spec, and both of them have been ramping up the power draw quite considerably in an attempt to beat each other. Both of them also want you to upgrade really, really badly, and now that we have both Ryzen 7000 and Intel 13th Gen on the table, we're going to discuss which one is looking like the better option. So I've got two of the new 13th Gen CPUs on hand here, the new 13900K which is the i9 flagship, and the i5-13600K which currently sits at the bottom of the stack. Now I'd forgive you for not keeping up with the absolute mountain of tech that's been launching recently, and for not knowing what's going on, so let's take a quick look at what is different here. The best way I think to think of these 13th Gen CPUs is a slight advancement over 12th Gen, with double the efficiency cores, a much higher clock speed, and a bit of extra cache. The new 13900K for example has 24 cores and 32 total threads, with a peak boost clock of 5.8GHz. Similar story when we look at the new i5-13600K, it's double the e-cores of the 12600K, now making it a 14-core part, with the boost clock now tipping over 5GHz. There is also a new i7, I won't be reviewing that one today, but here's a quick look at what you're working with, seems to fit pretty comfortably between the i5 and i9. So it's a big jump in the e-cores this time around, doubling the efficiency cores from 12th Gen is actually a pretty massive advancement, and the clock speed increase is kind of matching what AMD is putting out there as well with Ryzen 7000, but at least in the core count battle they are a bit ahead. Another positive against AMD, at least for this time around, is memory compatibility. The new 13th Gen CPUs fit the same socket as 12th Gen, and that also means that you can plug them into a cheap B660 or Z690 DDR4 motherboard and have a pretty decent value setup when paired with the new i5. You'll be spending a bit more on the other hand if you're going to try and pair the 7600X with DDR5, even with the new lineup of AMD B650 motherboards which are a decent bit more expensive than Intel's B660 boards. On the other hand, expect much longer platform support from AMD, about 4 CPU generations, whereas Intel generally only have motherboard support for 2. Now the point of platform support is one that people like to debate on quite a bit, whether it matters or not, and whether it's a point of going with Intel or AMD, but really only you know if that platform support is valuable to you or not, and whether you're going to upgrade every 2 or 3 CPU generations. The way that I see it though, this is a pretty important point if you're building a workstation system and a system that really benefits from upgrading CPUs every generation, a system that kind of pays you the faster it is really. If that's the kind of thing that you're doing, then yeah, the extra platform support from AM5 definitely is valuable compared to Intel, which would require you to swap out a pretty expensive motherboard every generation or two. As for gaming builds though, which might see a GPU upgrade every 2 or 3 generations and then a CPU upgrade every 4, 5 or maybe even 6 generations, especially considering how fast these CPUs are, it's a much less relevant point in my opinion. So with those things in mind, let's jump into the benchmarks and talk performance. The first thing that we should talk about and clear out of the way is the thermals and the power draw, specifically for the new 24-core i9-13900K. If you thought AMD's 7950X was a furnace, just wait until you see this thing. Here we're looking at Blender, and out of the box running stock settings on the new Asus Z790 Maximus Hero, the power draw peaks at 320 watts. That's the same as an RTX 3080. Now it can't maintain that sort of power draw even with a 360ml liquid cooler at full blast on an open bench, so it throttles back to keep itself from overheating. Now it only runs that unlimited power for 96 seconds, at which point it's configured to run back at 253 watts, which is a pretty similar power target to the 7950X. In the end, you're looking at all of the performance cores settling it at just under 5.2GHz. So yeah, 24 cores packed into that thing, it's gonna be pretty hot. But just like the 7950X, some quick BIOS tuning can fix that, dialing back the short and long duration power limit to 250 watts for example would be a pretty good idea. The 13600K on the other hand, really no issues there. The all-core clock speeds are tame enough, and the power draw seems to top out at only around 150 watts, and at that point you really don't need an extreme cooling solution. The 13900K though, very similar feeling to what AMD have done with the 7000 series. More cooling basically equals higher clock speeds and more performance. Speaking of which, the 13900K manages to just edge out the AMD 7950X in Cinebench, which considering how impressive that result is from AMD, that's a really surprising result from Intel. What's more surprising though is the new i5 beating the new Ryzen 7 7700X by a gap of 20%. These are just not the multi-threaded results that we're used to seeing. Then when we run on a single thread, the 13900K tops the chart here by a pretty healthy lead, 10% faster than the 7950X, whereas the i5 settles just behind them. When we switch to Blender though, that's where AMD has a few percentage point advantage over the new i9. Realistically, it's only a few seconds for this single frame, but if you're doing longer animations that can only render on the CPU, it is something to consider. And again, we see the 13600K beating the 7700X, which is actually more expensive. Next up, let's look at some video editing benchmarks. So here we have 3D motion tracking in After Effects, which is almost entirely a single threaded effect, and despite the 13900K having the fastest single threaded result in Cinebench, it's not as fast here as the 7950X. Same story for the 13600K, which realistically should be compared to the 7600X, but I just ran out of time testing that particular chip, so we'll get to testing that one in an upcoming video. Creating proxy footage in Adobe Media Encoder sees the i9 and Ryzen 9 pretty closely matched. Only a few seconds between them for what is a nice bulk of 4K footage, and it's important to note here that both what AMD and Intel have to offer is a nice leap over their previous generations. Unfortunately, I can't say the same thing about DaVinci Resolve. Here we're looking at export speeds where there really isn't much of a difference. Even if we wind back the clock to 3rd gen Ryzen, that's still keeping up pretty reasonably. But let's switch gears to gaming though. Maybe that's where we'll see a big difference between Intel 13th gen and AMD 7000. Well, not really. Here in Doom Eternal running 1080p with an RTX 3090 at almost 400fps, there's a few frames between them. Then, Forza Horizon 5, yeah, the differences are pretty sparse. In fact, throughout these gaming benchmarks, both Intel and AMD trade punches for who gets to sit at the top of the stack, and as far as I'm concerned, whether you go for Intel 13th gen or AMD Ryzen 7000, you have a top tier gaming CPU. Either CPU architecture will open up as much CPU bottlenecking as you'd ever hope for in your gaming system, and for that reason, it's not really worth me sitting here and picking apart the couple of percent here and there. Ryzen 7000 tops the chart in Valorant, but then it's Intel that leads in CSGO. And in either case, we're talking about both sides exceeding 700 frames per second. Now, I do think that Intel 13th gen chips, they are slightly faster when it comes to gaming overall versus Ryzen 7000, but it really is slightly. It's such a small difference, in fact, that if there are other features or qualities of the other platform that you might prefer, those probably hold more weight in your final decision than a few fps. Power draw might be one of those things. In Valorant, the 13900K pulled slightly more than the 7950X, and the 13600K surprisingly only pulled 75 watts on average. As we saw in Cinebench though, the new 13th gen chips have the potential to chug back a lot of juice, with the 13900K happily pulling over 300 watts like an absolute hippo. So the new 13900K is actually a bit of a monster compared to the 12900K, and it feels kind of similar to the 7950X, where it's just a giant leap in performance when it comes to production workloads over the 5950X. Here it's pretty much the same story with the 13900K over the 12900K. Where does that leave both of them? Well, honestly, pretty evenly matched. It's kind of insane how close the 7950X and 13900K are, both when it comes to production workloads in gaming, and it feels like for the first time ever that we're saying that. Like usually it's one or the other, like there's a slight trade-off for gaming performance versus multi-threaded performance, but here both kind of excel at both things. So look, if you're someone who sees yourself upgrading more frequently, whether you just really like upgrading or you have a workstation which can profit from being faster each generation, then the 7950X is probably what I'd go for. One reason to consider Intel on the other hand is if you need a lot of RAM. With the 13900K you can get a system up and running with 64 or 128 gigabytes of DDR4 for a fraction of the cost that you'd be forced to shell out for DDR5 paired with AMD. And that kind of fades into my recommendation when it comes to gaming, because if you're primarily interested for a CPU for gaming, the 13600K is actually looking really decent. You can pair it with a cheap B660 motherboard along with some DDR4 memory, and that means that you can focus a lot more of your budget on your GPU or maybe upgrading your monitor. Now all of my 13th gen testing has been completed with DDR5 memory, and pretty fast DDR5 memory at that, 6000 megahertz CL30, so stay tuned for a follow-up video where I additionally test the 13600K with a much cheaper DDR4 setup. In the end, it's pretty hard to say a win for Intel or a win for AMD. Both platforms have their advantages and disadvantages, but both are kind of landing in a very similar spot when it comes to actual performance. Both of them have just really accelerated the multi-threaded performance and production workloads and content creator workloads, whereas both of them are really struggling when it comes to gaming. It seems more about the gains of inches there rather than the leaps and bounds that we see with multi-threading. But I will leave them linked below, and I will be doing some more content on them soon, so stay tuned for that. As always, a huge thanks for watching, and I'll see you all in the next one.