i9 9900KS overclock result, good or bad?

[OCNewbee]

I recently procured a i9 9900KS. 

It is running at 5.0 core/ 0 avxoffset/ 4.7 uncore at 1.34v at LLC level 6 (asus z390i)

max temp is 92 and average is around 78 using occt.

*its been delidded 

*I am on a smaller air cooler (silverarrow 130) 

 

What do you think? Praiseworthy or nay?

[DoctorNick]

Nay. 

[Levent]

Average if not slightly above average. I would give it 6 out of 10.

[jaslion]

6 minutes ago, Levent said:

Average if not slightly above average. I would give it 6 out of 10.

On a sidenote here. Op has to run a cinebench at stock vs oc as it can very well be their oc performs worse especially with such a small oc.

[S.Hamed23]

20 minutes ago, OCNewbee said:

I recently procured a i9 9900KS. 

It is running at 5.0 core/ 0 avxoffset/ 4.7 uncore at 1.34v at LLC level 6 (asus z390i)

max temp is 92 and average is around 78.

*its been delidded 

*I am a smaller air cooler (silverarrow 130) 

 

What do you think? Praiseworthy or nay?

Very Very Bad for long run. As it is a KS cpu, this golden boy should run way cooler than regular ones.
Don't try to push beyond 1.35v. Set llc to auto, current limit at full, cpu cache at max (dial in 9999999), multicore enhancement to disable, avx offset to zero, sync all cores. When you have a base line, go for per core tweak. OC should not exceed 78-80C under long heavy load

[S.Hamed23]

@Levent what?! It's not a meme page

[Kilrah]

Nay, that's no better than my non-S 9900K does by just flipping MCE on in the mobo without any further tweaking, not even talking of the delidding...

[OCNewbee]

wow harsh audience.

 

you guys should really take into consideration that

 

1) I am on an ITX board. I think for ITX board, this is pretty good.

2) I am using a 130mm air cooler.

3) My avx offset is 0.

 

Considering all this, is this still that bad?

[OCNewbee]

10 minutes ago, S.Hamed23 said:

Very Very Bad for long run. As it is a KS cpu, this golden boy should run way cooler than regular ones.
Don't try to push beyond 1.35v. Set llc to auto, current limit at full, cpu cache at max (dial in 9999999), multicore enhancement to disable, avx offset to zero, sync all cores. When you have a base line, go for per core tweak. OC should not exceed 78-80C under long heavy load

what? CPU cache at max??? Is this a joke? beyond 4.7 cache, i9 9900k become very unstable. 

hmm multicore enhancement to disable? does this make a difference. Mine is on auto at the moment.

[S.Hamed23]

Just now, OCNewbee said:

what? CPU cache at max??? Is this a joke? beyond 4.7 cache, i9 9900k become very unstable. 

hmm multicore enhancement to disable? does this make a difference. Mine is on auto at the moment.

can cause crashes, my 9900k did. I dunno the settings specific to this motherboard, but mine related to the max tvb. In case of multicore enhancements kicks in if set to auto, that would compensate any possible crash. That's why I turn it off in oc

[Guest]

Think you have a respectable OC.

4000mt/s memory accommodates the cpu and cache oc pretty well. 

 

Nice work. 

Cpu will want colder temps for more though. 

[OCNewbee]

14 minutes ago, Kilrah said:

Nay, that's no better than my non-S 9900K does by just flipping MCE on in the mobo without any further tweaking, not even talking of the delidding...

I am curious, you are using a noctua l9i, which is tiny cooler that can barely handle stock temps. What is your overclock?

[jaslion]

16 minutes ago, OCNewbee said:

wow harsh audience.

 

you guys should really take into consideration that

 

1) I am on an ITX board. I think for ITX board, this is pretty good.

2) I am using a 130mm air cooler.

3) My avx offset is 0.

 

Considering all this, is this still that bad?

Have you tested it at stock in cinebench single and multi score vs your current oc? Sure you can have a 7ghz oc but if the score is lower than stock it is still a bad oc. I can't judge a oc without scoring data.

[Kilrah]

6 minutes ago, OCNewbee said:

I am curious, you are using a noctua l9i, which is tiny cooler that can barely handle stock temps. What is your overclock?

5 all core but it obviously can't hold that for long under the l9i, at sustained 100% load it'll have to go down to about 4.6. Still gives a boost for short duration loads/usage spikes which are most of what it sees in real life anyway.

I've had it in the other machine under the NH-D14 for a while where it was running at 5 all core sustained with no issues.

[OCNewbee]

8 minutes ago, jaslion said:

Have you tested it at stock in cinebench single and multi score vs your current oc? Sure you can have a 7ghz oc but if the score is lower than stock it is still a bad oc. I can't judge a oc without scoring data.

sorry man. just too tired to revert back to stock. 

Right now using cinebench r15, single core is at 208 and mutlicore is at 1952. 

[SFFDesigns]

15 minutes ago, OCNewbee said:

you are using a noctua l9i,

Nah, L9i is 37mm, he's using a 130mm cooler.

[Kilrah]

1 minute ago, SFFDesigns said:

Nah, L9i is 37mm, he's using a 130mm cooler.

You're mixing people up.

 

3 minutes ago, OCNewbee said:

Right now using cinebench r15, single core is at 208 and mutlicore is at 1952. 

2047 here on R15 multicore last time I checked, didn't do single

 

[jaslion]

3 minutes ago, OCNewbee said:

sorry man. just too tired to revert back to stock. 

Right now using cinebench r15, single core is at 208 and mutlicore is at 1952. 

Try a more modern cinebench. R15 isn't reliable enough for higher core count cpu's since it's from a time where quad cores were the norm.

What I can find is a stock 9900ks gives  about 220 single core and 2150 multi core in r15.

 

So yeah your oc as it is right now is performing worse than a stock 9900ks. So I would label it a bad oc.

[OCNewbee]

1 minute ago, jaslion said:

Try a more modern cinebench. R15 isn't reliable enough for higher core count cpu's since it's from a time where quad cores were the norm.

What I can find is a stock 9900ks gives  about 220 single core and 2150 multi core in r15.

 

So yeah your oc as it is right now is performing worse than a stock 9900ks. So I would label it a bad oc.

I have some bloatwares in the background. Asus armory crate and corsair icue. But at the same time, is it because I left multicore enhancement on auto?

[OCNewbee]

5 minutes ago, Kilrah said:

You're mixing people up.

 

2047 here on R15 multicore last time I checked, didn't do single

 

is your multicore enhancement on or off?/ nvm dumb question. its on obviously.

[jaslion]

3 minutes ago, OCNewbee said:

I have some bloatwares in the background. Asus armory crate and corsair icue. But at the same time, is it because I left multicore enhancement on auto?

All that really does is dynamic frequency. So no harm done. Close the bloatware and test again.

[OCNewbee]

4 minutes ago, jaslion said:

All that really does is dynamic frequency. So no harm done. Close the bloatware and test again.

2029. I couldn't close icue tho since I use commander pro.

https://www.eteknix.com/intel-core-i9-9900k-cinebench-15/

The normal according to this link is 2008~2166

[jaslion]

1 minute ago, OCNewbee said:

2029. I couldn't close icue tho since I use commander pro.

Margin of error basically. Still this is against a stock 9900ks so I do not think your current oc is well doing anything for better performance.

[OCNewbee]

Just now, jaslion said:

Margin of error basically. Still this is against a stock 9900ks so I do not think your current oc is well doing anything for better performance.

yeah but I didn't oc the core clock itself. The core clock is the same at 5. Cinebench r15 does not use avx. The two things that I overclocked are avx and cache. So it is unsurpising that the scores are the same as stock. I need to run a benchmark that uses avx.

[SFFDesigns]

10 hours ago, Kilrah said:

You're mixing people up.

 

Oh yeah I just realized that, whoops.