my 8700K initial OC results, is it pretty average?

newls1

newls1

Supreme [H]ardness
Joined
Sep 8, 2003
Messages
4,607
Using MSI Z370 PRO Carbon (non delided) 8700k using highend watercooling setup, and 32gb G.Skill Trident Z ram.... My results are:

5.04GHz @ 1.35 / 4.7Ghz Cache 1.3v / DDR4 @ 3605MHz 15/15/15/25 1.37v

This is "Realbench" 1hour stable (havent tried any longer as I feel that is more then enough) Highest coretemp reading was 68c and this is not even delided yet! My watercooling seems to really hold these temps well

so how did i do?
 
newls1 said:
Using MSI Z370 PRO Carbon (non delided) 8700k using highend watercooling setup, and 32gb G.Skill Trident Z ram.... My results are:

5.04GHz @ 1.35 / 4.7Ghz Cache 1.3v / DDR4 @ 3605MHz 15/15/15/25 1.37v

This is "Realbench" 1hour stable (havent tried any longer as I feel that is more then enough) Highest coretemp reading was 68c and this is not even delided yet! My watercooling seems to really hold these temps well

so how did i do?
Click to expand...


Way better than average.. very nice chip you got.. delided you probably can aim for the 5.2ghz mark..
 
i will test soon, have to get an actual day off again and a few hours to myself. once that happens, ill update this thread
 
Nice temps for not being de-lidded. I am at 1.360 @ 5Ghz delidded with around the same temps on RealBench with no offset. Latest prime95 is going to be a bit higher though.
 
Niiiiice! I haven't pushed mine yet, it is purring along at 4.9. I just want to hit 5ghz and claim I had been there... lol
 
Update: Just delidded the 8700k and increased the OC to 5.125Ghz @ 1.365v, with a 19c room temp and after running 15min of realbench, my highest temp was 58c. Mostly hovered at 52-53c.. Not a single core ever got over 58c. Not bad! I messed up and never did a screen capture, so either believe me or not, it happened LOL! I first ran this 5.125GHz @ 1.350v but after 11mins in the stress test, she BSOD'd on me. Increased the Vcore to where it is now @ 1.365v and Realbench completed... As soon as the stress test was done, I loaded up FC4 and played 30ish mins of that, and no issues. Im really liking this 8700k over the 1700x it replaced (for games)

EDIT*** Here is a screen cap of current temps near idle...

untitled751.png
 
Last edited:
Nice! I was able to pull 5.0 @1.30 (manual) + LLC6 with an AVX negative offset of 3 and all voltage limits set to Max. Basically der8auers YouTube guide ;-) , on custom water.

Mem: Corsair Dom plat 3200 2x16 with XMP
MB: Asus Z370 Code



With prime 95 using small, I hover in the high 60's to low 70's and sit there, 6 hours stable was good enough for me. Not too bad for a still lidded chip. Mine was an oem tray CPU from provantage.
 
Last edited:
Prime95, the newest versions only take 6-7 hours to get through every FFT set, so normally if you make it that long, youre solid! Nice chip!
 
I used 26.6, not the latest. Stayed away from the AVX instruction set, which ties into the negative offset of 3 in case I do need a heavy, %100 load of AVX (which I typically don't).
 
cybrnook said:
I used 26.6, not the latest. Stayed away from the AVX instruction set, which ties into the negative offset of 3 in case I do need a heavy, %100 load of AVX (which I typically don't).
Click to expand...

Yeah I rarely even use P95 anymore for stress testing.
 
I like using it once, at least during the initial OC to make sure it's stable. Tool sets I normally use are Cinebench, OCCT, Prime95, AIDA64 Stress test.
 
I use prime 95 small fft's (avx) in long cycles to produce the highest temperatures I will see. The reason I do this is to make sure that I can cool the chip under any load.
 
  • Like
Reactions: Meeho
like this
cybrnook said:
I used 26.6, not the latest. Stayed away from the AVX instruction set, which ties into the negative offset of 3 in case I do need a heavy, %100 load of AVX (which I typically don't).
Click to expand...
I guess I didn't realize the full impact of avx. On my recent 7700k I found that it's a must to test both avx and non avx as the vcore is bumped with avx. I was stuck at 4.7ghz avx stable with temps hitting 85c after hrs of p95 with avx and vcore of 1.3v.

I went back and played with non avx p95 and vcore was only 1.23v . So was able to run 4.8ghz(haven't tested much but sense then I bumped vcore to 1.28v and 4.9ghz is stable non avx also, hits 80c.

So I think the avx offset is useful, currently just running 4.8ghz with -1avx offset.
 
Last edited:
I do see AVX throttling kicking in during gaming and my typical day to day work loads. Although I don't specifically do things to purposefully cause it, seems it is still detected here and there. With my AVX negative offset of 3 meaning drop 300mhz when AVX is detected, it goes down to 4.7. For example, using afterburner (and Rivatuner really) I can watch my Core Clocks drop from 5000 to 4700 at times, then ramps back up to 5000. However, even under AVX loads my vcore stays at a solid 1.31/1.32 (due to LLC 6). When the loads dies off, my vcore drops back to 1.29~.
 
cybrnook said:
I do see AVX throttling kicking in during gaming and my typical day to day work loads. Although I don't specifically do things to purposefully cause it, seems it is still detected here and there. With my AVX negative offset of 3 meaning drop 300mhz when AVX is detected, it goes down to 4.7. For example, using afterburner (and Rivatuner really) I can watch my Core Clocks drop from 5000 to 4700 at times, then ramps back up to 5000. However, even under AVX loads my vcore stays at a solid 1.31/1.32 (due to LLC 6). When the loads dies off, my vcore drops back to 1.29~.
Click to expand...
So your running fixed voltage then right? Have you checked vcore running p95 with and without avx? Are they the same?

I'm on offset voltage, set to normal -.060v currently at 4.8ghz. All my speedsteps are enabled so it down clocks to 800mhz idle at .8v. My vcore is about 1.3v with avx(4.7ghz) and about 1.23v without avx(4.8ghz)

This may be from me using adaptive vcore.

I have seen other people say BF1 or PUBG may use avx, idk.
 
Yes, I am using "Manual" voltage (fixed). I started off the same as wanting to use adaptive like I was doing on my Z170 and Z270 chipset. But seems to not have all the gremlins sorted out on Z370 yet. I was noticing spikes of almost 1.5 booting into windows when using adaptive with an eventual turbo voltage set to 1.3, it was WAYYYYY overshooting. Setting to Manual/Fixed dialed it in right where I wanted it.

With that said, my clocks ramp down when not under load, but my vcore stays around 1.29. Of course as new BIOS updates come out, I will keep testing adaptive until it's sorted out. In the meantime I am not too worried since I am under water and my idle temps are in the low 30's.
 
cybrnook said:
Yes, I am using "Manual" voltage (fixed). I started off the same as wanting to use adaptive like I was doing on my Z170 and Z270 chipset. But seems to not have all the gremlins sorted out on Z370 yet. I was noticing spikes of almost 1.5 booting into windows when using adaptive with an eventual turbo voltage set to 1.3, it was WAYYYYY overshooting. Setting to Manual/Fixed dialed it in right where I wanted it.

With that said, my clocks ramp down when not under load, but my vcore stays around 1.29. Of course as new BIOS updates come out, I will keep testing adaptive until it's sorted
Click to expand...
My adaptive ramps vcore way up also set to normal, then ramps up again if I raise the LLC from normal.

So I had to stick LLC on normal, which seems to works well enough

I think if you lower your LLC your adaptive will work better. I also thought I read somewhere that llc6 on some boards ramps things up and llc5 was much better
 
Last edited:
Really need to get back in this- I was seeing full-on BSODs with PUBG despite being OCCT stable, and had to back down to 4.6GHz for all cores max...

Anyone else seeing latest PUBG with latest Nvidia drivers being a bit crashy with mild overclocks?
 
No crashes here @ 5Ghz but I am latest prime95 stable @ 3-4 hours w/ no AVX offset. I am using the drivers they released for pubg too, not sure if that's still the latest though.
 
Speedeu4ia said:
My adaptive ramps vcore way up also set to normal, then ramps up again if I raise the LLC from normal.

So I had to stick LLC on normal, which seems to works well enough

I think if you lower your LLC your adaptive will work better. I also thought I read somewhere that llc6 on some boards ramps things up and llc5 was much better
Click to expand...
I did actually play around with adaptive again, nagging me.

I got it as stable as my testing software would let me. I will try and list out my settings (for Asus Z370 Code) as best as I can remember, since I am not at home. Everything else at default/auto unless called out:

XMP for DDR3200 2 x 16GB, selected "No" on doing ALL adjustments, just loaded the profile.
AVX Negative offset 2 (Though I could likely raise this, or seemingly remove it back to 0)
Sync All Cores
50
LLC 5 (I am waffling between 5 and 6. At 5 with 1.31, I drop to 1.296 under load which runs fine. Then at 6 with 1.30, I move up to 1.32 under load, so a little bit of an over shoot with some small heat increase. If 5 proves unstable, I will drop vcore to 1.30 and use 6 since that was 6 hours OCCT with AVX stable).
Long Duration PPL - 4095
Short Duration PPL - 4095
IA AC Load Line - 0.01 (These are what I needed to tame my 1.5 overshoot when booting into desktop)
IA DC Load Line - 0.01 (These are what I needed to tame my 1.5 overshoot when booting into desktop)
CPU Core/Cache Current Limit - 255.50
Core voltage - Adaptive
Eventual Turbo Voltage - 1.31 with LLC 5 (1.30 with LLC 6)
 
Last edited:
You must log in or register to reply here.
Back
Top