07-08-2018
08:35 AM
- last edited on
03-06-2024
08:23 PM
by
ROGBot
07-26-2018 02:17 PM
Phillyflyer wrote:
Finally managed to get the 8086k overclocked to 5.3 ghz and prime 95 stable with an 8 hour run. using bios 1503. Chip has been dellided with a copper heatsink and ambient temp at the time was 22 celcius. Vcore was set @ 1.425V with LLC 6 with the GSkill Trident Z 3866 MHZ running 1.35v with a VCCIO set @ 1.15v and VCCSA set @ 1.15v .
07-26-2018 04:06 PM
ftln46 wrote:
P95 with AVX or without ?
07-27-2018 05:07 AM
dmt.taran wrote:
Yea.....my 8700K does only 5.1 GHz on that Vcore.
P.S. Isn't that a little high for 24/7? I would run this one at 5.0@1.2, if I had one
07-27-2018 05:47 AM
Phillyflyer wrote:
Had the 8700k previous to the 8086k....chip was P95 solid @ 5.0ghz @ 1.28v (1.296v under load) and @ 5.1ghz @ 1.36v (1.376v under load)....5.2ghz seems to be the wall and needed upwards of 1.435v ++ to be stable.
As for 5.0ghz @ 1.2 v ...not attainable...have yet to see a chip that can undervolt that low @ that frequency and maintain P95 stability.....may work with Aida64 but not true stability.
Yes....may be a tad high for 24/7....according to Intel spec chip can volt to 1.52v though....currently have it @ 5.2ghz @ 1.375v (1.392v underload) 24/7
07-27-2018 06:24 AM
dmt.taran wrote:
https://www.hardwareluxx.de/community/f139/intel-coffee-lake-s-sockel-1151-oc-ergebnis-thread-kein-q...
Dunno, I doubt those results are truth (or it's just boot voltages, not 100% stable) - the best seems like can take 5.0 at 1.15V. Won't beleive it until see for myself though 😄
As for me, I wasn't lucky to win the lottery - mine can do 5 GHz with LLC5/6 only at 1.36V. The fact that it requires >1.35V and LLC obligatory says that silicon isn't so good. 😞
07-27-2018 07:30 AM
Phillyflyer wrote:
Bro....I tend to pick things apart when I see screenshots like that....went through a few screenshots in your link....and in the 5 mins I breezed through them...the 1st thing that stood out is that the majority only ran P95 for 45 - 60 mins.....I'm like you...I'd have to see it to believe it !!
07-27-2018 08:14 AM
NarfTV wrote:
Yup that are the rules for the listing in the hardwareluxx forums. It is described in the threads description. Just to make sure everyone is running the stability test under same conditions to be listed.
Its not really that hard to get such a listing and it doesnt 100% reliably proof stability. But be sure that 8700k's @1.168v vcore 5GHz stable do exist. Ive seen some going through the whole prime95 23h test. But ill mind you that this is a vcore reading when using Asrock Boards. They tend to have a different sensor placement showing less vcore and also setting more vcore than shown when typing it into the Bios but its just a notch.
Rest is a matter of chip quality and cooling solution. Its a huge difference if you f.e. use a custom loop with a Mora or an AiO.
07-28-2018 10:48 AM
07-29-2018 04:25 PM
Phillyflyer wrote:
Welcome my friend....was actually wondering if you would chime in.....Yep...there may be rules...but we all know there is away around to manipulate that...lol......As you mentioned... I did notice that the bios vcore and and what was shown to be different...bios vcore actually higher.....With the Asus board LLC 5 would do that with a bigger drop....LL6 would have increased it...….The screenshots that I looked all were at LL6.
This is what really got my attention though:
https://abload.de/image.php?img=karhu4266gqf14.png
Then this:
https://abload.de/image.php?img=5400w1sdc.png
The latter apparently running for 3 hours P95 yet no screenie
Care to share the settings ??….Cuz I have more questions....as my interest is peaked.
This is what I have noticed from the first screenshot:
The Aida64 Benchmark depicts the cpu clock @ 4000...the cpu multiplier @ 40x...the northbridge @ 40 and the board as an Asus ROG Strix Z370-G Gaming...now...CPU-Z depicts the chip is clocked @ 5.4 ghz with the northbridge @ 50 with a GSkill Trident Z RGB memory kit F4-3866C18 ?? - 32GTZR ( I would assume a dual kit 2x16) that has been overclocked to 4266mhz. The screenshot was captured on 23/7/2018 @ 23:13pm
The 2nd screenshot depicts the same chip...same board same northbridge frequency but was captured on 20/06/2018 @ 10:13am.....funny thing about the 2nd screenshot is that is when it was posted in this thread entitled( 8086k - Z370-i - 4000mhz Ram) on 07/24/2018 it was stated to be still running Prime 95 for 3 hrs ....Look close @ the P95 date... June 20 /2018.
Even more interesting is that in this thread on 07/18/2018 entitled ( Cant get G.skill kit working on Z370 Maximus X Formula ) the board that was mentioned was an Asus z370 Maximus X Formula ….the chip 8700 K ….and memory GSkill F4-3600C15D-16GTZ memory and the following quote:
07-20-2018 05:46 PM
#55
NarfTV
New ROGer
Array
Join Date
Apr 2018
Reputation
10
Posts
45
Quote:
"yup, thats also a reason why im a little mad. I mean i could have just taken the 3600cl16 RGB kit. RGB would give me the performance boost that i need. Just kidding.. well yeah the problem is just no matter what timings/bandwidth i set. If its faster than 8.5ns its not gonna get stable. tried 4133CL17 for some hours now. Cant get it stable either.
Maybe silent scone is right, and its just the IMC of my CPU. Could be crappy though i dont know. "
And on 07/24/2018 ...Quote
You know, Karhu is still running for 3hours at the screenshot ive posted . Do you think there is even a slight chance that this CPU could be unstable? NO there isnt, should be proof enough that it is his 24/7 OC and fully stable. You just have a worse chip than him. Infact your chip is even worse than mine.
The proof that you want so hard:
https://abload.de/image.php?img=5400w1sdc.png
Need I say more as to the validity of those screenshots ??