i5-12600K vs Ryzen 5 5600X with an RTX 3060 for GAMING
The vast majority of reviews and benchmarks you’ll have seen with these new Alder Lake CPUs, including this i5-12600K, have been with top shelf GPUs like an RTX 3080. Even my reviews were all with an RTX 3080, but since no one bar Bezos can get their hands on one of those right now, I wanted to see how this stacks up with a more ‘mid tier’ card like this RTX 3060 – which yes I know is still selling for top tier flagship money, but still.
Since the CPU specific results won’t change with this new GPU, I’m sticking exclusively to gaming here. If you want to see how these chips stack up, including to a few other options, do check out the full review I’ll leave in the cards above for you! I’m testing at both 1080p and 1440p here so you can see how they stack up there, and I’m testing at what I’d call “reasonable” settings in game. Rather than using the highest, “ultra”, settings which often needlessly hurt performance for next to no visual differences, I’m testing at the sort of settings I would actually play at. For CSGO for example, that’s with everything pretty much set to lowest. Watchdogs Legion is on medium, Fortnite is on High, Shadow of the Tomb Raider is on medium, Microsoft flight is on medium too, and Cyberpunk is on medium with high textures. I’ve disabled DLSS in the games that support it, although if you’d like to see some more tests with that enabled feel free to let me know.
Ok, let’s get to the games! Let’s start with the oldest game, CS:GO. On low settings, at 1080p, the 5600X actually takes the lead here with over 20% more performance on average netting 536FPS average, up from 441 FPS on the i5. This is the same pattern I saw in my review testing with both the i9 and i5, and in this case even the 1% low numbers are around 5% up on the i5 too. At 1440p it’s a similar gap, this time more like 18% faster, netting just over 500 FPS versus 430 FPS, and again significantly improved 1% low numbers too.
Watchdogs Legion swings the other way in the 12600K’s favour, netting 119 FPS average at 1080p, up from 107 FPS on the 5600X. The 1% lows are also up, at 94 FPS on the i5 and 80 FPS on the Ryzen 5. Unsurprisingly though, at 1440p there is much less of a difference, just 2 FPS difference with the i5 still just about holding the lead.
In Cyberpunk at 1080p the i5 holds around a 7% lead, averaging 94 FPS up from 88 FPS on the 5600X. That’s a pretty close call, still a win for Intel here obviously but would be pretty hard to notice in game. That is, if the 1% low figures weren’t so starkly different. The Ryzen 5 nets just under 60 FPS, whereas the i5 nets more like 73 FPS. Of course that’s hardly game-breaking, but it’s a 25% advantage to Intel there. Interestingly at 1440p the 5600X actually takes the lead, but really that’s within margin of error at just a 2 FPS difference. The 1% lows are equally tied, although the i5 is the one in the lead there.
Fortnite shows very little performance difference between the CPUs on high settings as both chips remain within 1 or 2 FPS of each other, on 1080p and 1440p, average and 1% lows. Technically the i5 is in the lead on all counts, but with such a marginal difference you will not see or feel that difference at all.
Microsoft flight is really neck and neck for both 1080p and 1440p, and on average and in the 1% lows, with Ryzen taking the win at 1080p by a single FPS average, and Intel ahead by 2 FPS at 1440p. I think it’s fair to say no matter which chip you play on here, you are going to be more GPU bound than anything.
Speaking of GPU bound, I’ve added Shadow of the Tomb Raider to my benchmark list now so I can show you those results! Both at 1080p and 1440p both chips scored identically, at least in the final FPS actually displayed. Shadow of the Tomb Raider’s built-in benchmark is rather interesting though, as it includes a lot more information than you’d normally get from running benchmarks manually. Specifically, it lists how GPU bound your run was, which for both chips at 1080p was 99% and at 1440p was 100%, but it also gives you a little table of the min, max, average and 95% figures for both the CPU Game, CPU render and GPU during the run and if we plot those results, the CPU render results, you’ll see the i5 offers significantly more performance. It’s consistently around 23% faster on all counts, smashing past 300FPS average on both 1080p and 1440p whereas the 5600X is more like 250 FPS. Not absolutely insane, but shows the raw performance of the i5 even if that doesn’t translate to any difference in your experience.
At this point I think it’s worth making clear that these results were captured on Windows 10, with the most recent BIOS version (0702) on an Asus Z690 Formula board with Corsair DDR5 (although not running with XMP enabled due to what seems like a bug in the BIOS causing lower performance with both memory frequency and timings set, even manually), and most importantly with all 4 E cores disabled in the BIOS. There is one very good reason for that, which is that a collection of games with various anti-cheat or DRM solutions flat out refuse to run with the E cores active. In my testing, that was Watchdogs Legion. It would show the splash screen, sometimes even show a black screen, then close with no error messages. After trying literally everything including uninstalling both the game and Ubisoft Connect, and wiping every trace of both from the system then reinstalling both from scratch, nothing worked. Then I disabled the E cores and as if by magic it ran just fine. That is something you might need to consider if you want these new Alder Lake chips.
With that said, on the whole the i5 should be a faster chip – that’s illustrated nicely in Tomb Raider’s extra numbers – but in practice at least in these games, in my testing, it’s actually a hair slower than the 5600X when paired with this RTX 3060. In fact if you average every 1080p average FPS result the i5 and Ryzen 5 each got, you’ll find the Ryzen chip is around 7% faster, and 8% at 1440p. Only when you look at the 1080p 1% low figures do you see the i5 coming out on top at around 4% higher. That should equate to a more stable experience while gaming, but we are talking about a 4 FPS difference here so it’s hardly substantial.
It is also worth noting the price difference – of your whole purchase. The CPU itself is actually currently £10 less than the 5600X at least on OverclockersUK – although on other sites the 12600k is either equal or more expensive – but you’ll need a Z690 motherboard to use this chip, which will set you back at least £200 for a reasonable DDR4 board versus more like £120-150 for a B550 board, and you might need a better CPU cooler too. Plus if you want DDR5, well currently good luck getting any and even if you could it’s likely to be triple the price of DDR4. If you compare the cost of the CPU, RAM and motherboard I specifically used here, while not exactly a fair comparison as I’m not using the same class of board, the AMD parts would currently set you back around £550, and the Intel ones more like £1200. Even with a much cheaper motherboard, you would still be looking at around £800, so it’s well worth keeping that in mind at least until B660 boards are available and hopefully cheaper.
