I9 9900k

[moorhen2]

Iirc i'm at 1.46v to get r15 5.4ghz to pass. But having to cool water a little, idle temps circa 15c and 80-85c under load

Nice, but you are using a bucket and pump with ice cold water I believe.

 

[Nickolp1974]

Nice, but you are using a bucket and pump with ice cold water I believe.

Yeah water will be about 6-7c can't go colder as board Isn't insulated, that can wait till I get back off holiday then I can go colder.

 

[Majek]

Majek, sorry if we have hijacked your thread. Apologies.

Oops, its not your thread, sorry again.

Hi, The OCN 'losers' thread is mine. No worries.

Also, agreed on reporting voltages. It is hard to get 100% fair comparison with different boards, coolers, cases, llc settings, ambient etc.

In my case it is further exacerbated by the fact that Aorus Master has two sensors and one of them gives much higher vcore reading than the other one. Apparently, the one to look at is IT8692E as it is closer to what a multimeter reads.

Thanks,

 

[Distracted]

Nice, but you are using a bucket and pump with ice cold water I believe.

Yeah water will be about 6-7c can't go colder as board Isn't insulated, that can wait till I get back off holiday then I can go colder.

This makes me chuckle. So ghetto mod.

 

[moorhen2]

Just finished a 1.5 hour Aida stress test at 5.2ghz core and 4.5gh cache. Highest temp 79c, average full load voltage 1.279v. Next up cache @46.

 

[0121danwilliams84]

Just finished a 1.5 hour Aida stress test at 5.2ghz core and 4.5gh cache. Highest temp 79c, average full load voltage 1.279v. Next up cache @46.

this is exhaustive testing!

 

[moorhen2]

Tell me about it, but I do enjoy tinkering lol.

 

[0121danwilliams84]

this is exhaustive testing!

curious about what temp 9 is? VRM's?

 

[moorhen2]

curious about what temp 9 is? VRM's?

Don't worry, its an unconnected sensor, will be removed from the next build of HWiNFO64, and no not the VRM's, they are not monitored.

 

[Majek]

That's a good idea. I guess as time goes by we will see what the average results are. I know it won't make much difference to my experience but it's always nice to get the extra performance and it looks good at 5GHz too! I found the losers thread! What motherboard are you using? I am using more voltage than you 1.340v for 4.9GHz not 5! Thanks.

Hi again,

Just FYI - I thought I was stable with uncore set to 43 and voltage at 1.305 but I am not. Prime 29.5 just gave me a WHEA. I need full 1.32 it seems.
May I ask what uncore/cache ratio you have when running yours at 4.9/1.34?

Thanks!

 

[0121danwilliams84]

Hi again,

Just FYI - I thought I was stable with uncore set to 43 and voltage at 1.305 but I am not. Prime 29.5 just gave me a WHEA. I need full 1.32 it seems.
May I ask what uncore/cache ratio you have when running yours at 4.9/1.34?

Thanks!

Hey, I am running cache on auto. Was messing with my adaptive voltage on 43.

 

[moorhen2]

Hi again,

Just FYI - I thought I was stable with uncore set to 43 and voltage at 1.305 but I am not. Prime 29.5 just gave me a WHEA. I need full 1.32 it seems.
May I ask what uncore/cache ratio you have when running yours at 4.9/1.34?

Thanks!

Why may I ask are you using P95 to stress test ?

 

[Majek]

Hey, I am running cache on auto. Was messing with my adaptive voltage on 43.

Ok, thank you! I am trying 43 now too.

 

[Majek]

Why may I ask are you using P95 to stress test ?

Hi,

the honest answer is that it seems most other people use it. Also, my experience of oc with current boards/cpus is limited, obviously, so I am just checking what the cpu is capable of and learning along the way.

What would you recommend instead? Have tried occt, prime avx and non-avx and CB15.

Thanks!

 

[moorhen2]

Hi,

the honest answer is that it seems most other people use it. Also, my experience of oc with current boards/cpus is limited, obviously, so I am just checking what the cpu is capable of and learning along the way.

What would you recommend instead? Have tried occt, prime avx and non-avx and CB15.

Thanks!

P95 has not been regarded as the best stress test for Intel chips for some time now, I have been overclocking for years now, and have never entertained P95, even though there is a non AVX version available. JMO

 

[0121danwilliams84]

Anyone notice any differences in temp reporting software? I get max temp 91c in HW info compared to 87c in Core temp. Maybe just quick spikes (Intel are known for) not caught at the same time. There's also massive difference in core temps. One barely gets to 80 and others more towards 90 max.

 

[moorhen2]

Anyone notice any differences in temp reporting software? I get max temp 91c in HW info compared to 87c in Core temp. Maybe just quick spikes (Intel are known for) not caught at the same time. There's also massive difference in core temps. One barely gets to 80 and others more towards 90 max.

I tend to use HWiNFO for more accurate temp measurements, Core Temp seems a bit off to me. As to core temps you will always have variations between cores, I sometimes get a 10c difference between the highest and lowest temp, but it is normal.

 

[Distracted]

I tend to go by HWinfo because it reports the highest temps for me. I also see differences between cores as well as spikes, which I understand is normal.

 

[0121danwilliams84]

I tend to go by HWinfo because it reports the highest temps for me. I also see differences between cores as well as spikes, which I understand is normal.

I am finding that games are usually in the high 60's and 70's but I get random 1 second spikes into the high 80's. Average temps over an extended heavy game session are about 70c. Seem about right? or a bit on the warm side?

 

[moorhen2]

Skipped the 46 and 47 Cache multipliers, straight to 48. Image should show an hours run of Aida64 core and cache stress test, but cant get the image to work.