i7-13700K vs i9-12900K Max Overclocked (Ambient) Gaming and Synthetic Benchmarks

[@rirochino]

Realy good video

[@pm110978]

Worth upgrading from a 9900k to a 13700k?

[@EraaCOD]

I would say yes

[@pm110978]

@@EraaCOD Or would the 13900k be a better upgrade option?

[@dainluke]

@@pm110978 The 13900K is a fine CPU if you actually do a lot of rendering work or if you do XOCing/3DMark. If you only game then avoid the 13900K. The extra E-cores are just a waste of money and extra voltage/power/heat by that point. The 9900K is still totally fine if you game at 4K, ultrawide, or 1440p max settings or using ray tracing. If you play stuff like Call of Duty or any game that runs at 200 FPS regularly, then the jump from Coffee Lake to Raptor Lake will definitely be noticeable. I’ve noticed that used Z390 and used 9900K still sell for a decent amount. If you can pay a minimal amount to upgrade then definitely go for it (if you want/need the frames).

[@myteor5576]

Hey man what kinda uplift we looking at with ddr4 to ddr5 ?

[@dainluke]

I haven’t had a D4 board to test this unfortunately. All hardware I test with is out of pocket so I’d need to buy one. I actually might do that though, I kinda have a source for one. I would wager that it’s not going to end up being too big though.

[@myteor5576]

@@dainluke Yeah that's what I figure to maybe higher in the 1% lows than the averages tho btw great channel dude trying to go through it haha 👍🏽👍🏽

[@dainluke]

@@myteor5576 It’s entirely possible, but I’m not sure. It’s like, at 4000-4100 C15/16 depending on where this chip maxes, latency might be 46ns in AIDA vs 49ns with 7200C32. I don’t think 3ns would be enough to affect 1% lows when bandwidth on 7200 is nearly 2X and the L2 is so much higher on RPL. Bandwidth of 112GB/s vs something like 65-70 on the best possible D4 tune means you’re moving a lot more data every time the IMC sends a data command. Combined with not that much slower latency, it starts to not really look good for D4 overall. I guess until I test it and compare it to this batch of results it can’t be made certain. Unfortunately only two of the games display 1% low numbers which is why I didn’t bother including them. 5/7 don’t for some reason. If you can spot them in the raw testing video then you have them though, I just could not spot them in the other games aside from Legion and Valhalla.

[@dainluke]

@@myteor5576 Well, it turns out I have a D4 platform now. Will test later when I get back from my exam. It’s a Strix-A and some decently binned 2x16 B-die. Will see what it can do (and my chip SP).

[@ashereu]

isnt win10 better for performance than win11

[@dainluke]

Honestly, no idea if that ever even was the case

[@shoxruxbekabduraxmonov6958]

Is 13700K better than 12900k in all tasks?

[@dainluke]

Not AVX512, because it doesn't have 512

[@shoxruxbekabduraxmonov6958]

@@dainluke 12900k also doesn't have avx512. Firstly 12gen intels had after 10th November they removed it with microcode

[@dainluke]

@@shoxruxbekabduraxmonov6958 I guess that's true, but I've never handled one without it. I'd say if you were buying one you should try to avoid those without because it's an objectively worse CPU

[@shoxruxbekabduraxmonov6958]

@@dainluke i think it had better to buy 13700k even with avx2 it would be better than 12900k

[@dainluke]

@@shoxruxbekabduraxmonov6958 yeah I'd definitely buy the 13700K