Intel i9 12900K – World of Warcraft Benchmarks + Memory Scaling



Read more about Shadowlands ➜ https://worldofwarcraft.mgn.gg

Intel with the 12th Gen CPUs is now the best performer in World of Warcraft Shadowlands, in this video you will see the 12900K with a lot of different tuning profiles, a complete memory scaling and overclocking deep dive to make this CPU shine at his best.

Timestamp:
0:00 Intro and expectations
0:43 CPUs @ XMP
1:16 CPUs @ XMP – E-Cores OFF
2:38 CPUs @ Fully Tuned RAM
4:30 Memory scaling SMA Off
5:09 Memory scaling SMA ON
5:47 Memory scaling 6900XT
6:22 Oribos Gameplay
7:30 BG 40vs40 Gameplay
8:27 Thoughts and conclusion

Base system:

Asus ROG Maximus Apex Z490, Z590, Z690
Asus ROG Crosshair XIII
AMD Ryzen 9 5950X
AMD Ryzen 7 5800X
GSkill Trident Z 4600 DDR4
Gskill Trident Z 6400 DDR5
ROG Thor 1200W
EVGA RTX 3090 Kingpin (XOC Bios)
PowerColor Radeon RX 6900XT Liquid Devil Ultimate
Alphacool Eisaber Extreme
Windows 11

_____________________________________________

If you need help to push the maximum performance out of your system join my Discord server here:

https://discord.gg/7pqpWNQ

If you want to support my channel you can shop in Amazon using the link below:

https://www.amazon.com/shop/hardwarenumb3rs

Or be a Patreon supporter:

https://www.patreon.com/Hardware_Numb3rs

____________________________________________
Follow me on Facebook

https://www.facebook.com/HardwareNumb3rs

Follow me on Twitter

https://twitter.com/HardwareNumb3rs

Follow me on Instagram

https://www.instagram.com/hardware_numb3rs

My Twitch Channel: https://www.twitch.tv/hardware_numb3rs

My HWBot Profile: http://hwbot.org/user/hardware_numb3rs/

source

49 thoughts on “Intel i9 12900K – World of Warcraft Benchmarks + Memory Scaling”

  1. Been a while since I've had an intel product let alone a laptop, but I gotta say…I've been digging my i7 Alderlake Laptop. Boost pretty nicely on the P cores to 4.6ghz. Really stoked that intel is swinging back now.

    Although I wish I could fiddle with it as easily as I could with desktop parts. Been having a ball with my 5900X.

    Reply
  2. Thus far, I haven't found a good reason to disable e-cores on my 12700k –
    The difference between maximum achievable Ring Frequency* with e-cores enabled and E-Core L2 Voltage around 1.3v, versus e-cores disabled entirely, is 4.5 Ghz vs 4.8 Ghz

    While that seems like a nice jump, in practice the difference ends up being roughly 0.9-2.1% – as opposed to the 4-6% you might hope for.
    And there is a reason for that – the e-core cluster has its own separate L2 cache, and if you force most background tasks onto them, you essentially free up 2 MB (or 4 MB on a 12900k) of L3 cache from being touched by background junk.

    That, combined with less context switching between P-cores to handle background noise, ends up reducing the difference to those ~1.3%

    Forcing a lot of junk onto the e-cores is easily achievable using Process Lasso, and unlike mystery people on the internet, I have not had a single issue in doing so.
    Perhaps it's that I use batch files with things like "start /affinity FFFF Overwatch.exe" instead of relying upon the program to set my games to not touch e-cores?

    * as for "maximum achievable Ring Frequency", then currently (? I don't update my BIOS when it already works, sorry🤡?) a Ring multiplier past 40x is treated more as a "suggestion" than an order. You can still figure out what your maximum is, but you won't actually see performance gains from it.
    HOWEVER as it turns out, you can simply use BCLK and a Ring multiplier of 40x to hit your Ring ceiling – and see the improvements suddenly rush in!

    Of course, that presents its own challenges with having your memory overclock line up with BCLK 112.6 or 120 and be perfectly (day-to-day, non-month-long-science-calculation) stable..
    But hey, at least it's not watching paint dry for a few hours with Vermeer and its SOC/IOD/CCD +/- 0.01v voltage dial-in process, heh heh heh

    by the way discord bannaned my 7 year old account and i don't feel like coming back :- )

    Reply
  3. super test.. thank you.. Personally dont like to turn features off with the gear i buy.. like turning off e-cores.. Zen 4 might be the last cpu without these small cores.. might be the one to go for with fast ddr5 memory..

    Reply
  4. Winter is coming and so is zen 4 rip alderlake lol

    Also it's probably just WoW getting massive benefits with disabling ecore because other channels like Hardware Unboxed showed with a 30 game benchmark that it mostly hurt performance in a wide verity of games. I'm guessing blizzard just needs to do an update and you should show them your findings! Great review…also i9 isn't worth the price at all, the golden chip for play and work is the i7 12700k, it also doesn't hit 100c like the i9 does in rendering tasks lol, 73c in heavy rendering and 34-42c in gaming, but if you only game the 12400f -12600k is where it's at

    Reply
  5. Thanks max, love your videos!
    Think you'll make a video on 5000X3D chips? Now that my B350 board has a bios update for 5000 series chips i need to decide on 5800 or go for a X3D version.
    Edit: I see you answered this already, looking forward to it!

    Reply
  6. Hey I have a question. I just bought an alienware x17 laptop and it has 4k res. 11th Gen I9 / 3080 and 32g ram. I get like 30-40fps.

    I set it to the default OC and it still has really low fast. It has the command center for tweaking boost settings. What should I put it at to get a solid 60-70fps and not ha e the pc crash?? Any help would be great thanks

    Reply
  7. Do you think overclocking alone is worth it vs stock? Currently running 5.1Ghz P cores and 4Ghz E cores on my 12700K. I'm just not sure if the extra performance is worth the extra watts.

    Reply
  8. Could you test the 5800X3D vs a normal 5800X when overclocked? Maybe per core curve optimizer, high fclk, fmax_offset and some scaler?

    Maybe even both against the 12900k OC or 12900KS (should be the same).

    Reply
  9. A fully Tuned good B die ddr4 kit with a good 12900k would be good to look at too, the ddr4 kits are a lot cheaper and so is the motherboards. Not to mention ppl with current ddr4 systems can just use their current kit. When overclocking the cache, cpu and memory the ddr5 with the new "b-die ddr5" hynix chips only has a slight edge compared to bdie ddr4. Would have been good to see what you could get with it.

    Reply
  10. I have a 6700k with CORSAIR Vengeance DDR 4 3600 C16. I’ve tried following some guides but most of the videos out there have newer CPU’s. Wow is my main game and I wouldn’t mind getting more FPS out of it. I run the XMP but I think the board limits me to 3200, I can’t remember without looking at it again. Thanks mate!

    Reply
  11. The E cores are a joke. Intel fucked up, it's all a marketing lie. They lower performance and they are not efficient either. I looked at laptop 12th gen reviews and it's hillarious. They crawled back to AMD's level on the performance side, but the efficiency is a bit worse than 11th gen.
    As i said in another channel's video, the CPU work to complete a task, a bigger core will spike to finish the task very fast then go back to sleep, when a smaller core will have to work for longer. The average power consumption will be the same, the smaller core will be cooler but slower.
    The need to disable useless cores is pathetic on the intel's side.

    Reply
  12. Came from Moore's Law is Dead. I'm someone who was never recommended your channel before, but I was awoken to memory overclocking with Jufes from FrameChasers. So refreshing to see someone else in the techspace doing memory overclocks with their CPUs. Earnt yourself a new sub!!! (now to watch the MLID interview… haha!)

    Reply
  13. Literally THE BEST channel testing hardware vs WOW! I always come here when looking for how stuff fairs out in wow as I mainly play that, when work allows me some time. Thank You!

    Reply
  14. Got a question. Recently updated my rig to a z690f and Intel i7 13700k. With that setup I have micro stuttering respectively frametime spikes in WoW in places like Oribos or Orgrimmar. My suspicion is that WoW kind of uses the e-cores instead of p-cores. Shiuld I disable them e-cores?

    Reply
  15. I know this is a old review but i want to know if the Ryzen 7 5800x 3d is still the best for gaming only I don't care about any thing else a pc can do If you could massage me or some one but I'm looking to get rid of my 12 core Ryzen 9 3900x Because amd came out with The best gaming CPU every the Ryzen 7 5800x 3d A year and a half after I build my pc and a upgrade would only be 400 dollars

    Reply
  16. Someone only just linked me this video after MONTHS of the constant arguments about tuned DDR4 vs tuned DDR5. All the major tech channels just hit XMP and when they try to test 4000CL14, its in gear 2.

    Thank you for properly testing the ram, maybe unfortunately a bit too late for me as I kept my discounted micron kit that did 4600CL15 on 10900K, currently doing 4300CL14 G1 on 13600K. I bought the kit on sale when DDR5 launched, what I always do, also just picked up a 12100 regular with iGPU on sale for £116 vs £149 for a 13100 for my test bench update.

    Reply
  17. Great video! Wondering maybe could offering what cpu upgrade,because now running rtx3060 ti i5 10400f ddr4 32mhz 16gb samsung evo 980 m.2 and still having shatering on dragonflight, have read some forums and it looks like its common problem,even lowering graphic getting same shuterring,windows all optimize but it looks like something is with cpu…

    Reply
  18. Very nice video! I like that you test WoW CPU scaling since MMOs in general are usually very CPU bound due to the network overhead (especially with a lot of people around like raids and battlegrounds). Also I wish you could to do a DDR5 memory tuning guide for 13900K.

    Reply

Leave a Comment