Stop me before I break something!

[Lwizzit]

5 minutes ago, Cracklingice said:

I use MSI Afterburner with Rivatuner Statistics Server.
As far as shocked by the temps, Intel dropped the ball big time with the TIM on seventh gen K skus.  Replacing with liquid metal sees drops in temps that are outright shocking to me.
There's an intel software iirc that will show you actual individual core temps in addition to the package temp.  Package temp can easily be 10 degrees lower than core temps.  Core temps are more important in knowing the actual temp of the silicon as a result and why when using package temp you should shoot for a cooler temperature as you don't actually know the core temps.

Yeah - I have been keeping an eye on the individual cores. With one exception the core temps were always equal or lower than what I have been reporting on the package.

But, following your advice, I have dropped Vcore to 1.20 - had to lower to 47X. That's fine.

Now the max temp I am seeing during Cinebench run is 76 degrees (core 1 - always seems to be the hottest). Package as well is reporting 76 max.

 

I am seeing 33-35 right now, with odd spiking to mid 50s.

 

I am going to hold this setting for now and continue testing.

 

What about in-game temperature reporting? Can you suggest something for that? I have Ghost recon - seems like a pretty good program to push everything hard....

[Cracklingice]

1 minute ago, Lwizzit said:

What about in-game temperature reporting? Can you suggest something for that? I have Ghost recon - seems like a pretty good program to push everything hard....

 

12 minutes ago, Cracklingice said:

I use MSI Afterburner with Rivatuner Statistics Server.

 

[Lwizzit]

Just now, Cracklingice said:

 

 

OK - I have that. Thanks again for the help!

 

You figure I'm OK for now with these settings and temps?

[Cracklingice]

23 minutes ago, Lwizzit said:

OK - I have that. Thanks again for the help!

 

You figure I'm OK for now with these settings and temps?

I'd say you're in a safe zone.  You may be able to push more, but I'm not the person to recommend it.  I think you would likely see great results in temp from a delid as I've never seen poor results from doing so on a 7th gen, but it is quite risky.  I've seen many examples of a 20°C drop with delid and liquid metal tim.
I'd also consider checking out the fan and pump speed on your AIO.  It seems like your temps are a bit high for the voltage you're stating.  It could also be that you're stating what you set it at and not what it is reported at during a stress test as the voltage does change with line level control.

[Lwizzit]

23 minutes ago, Cracklingice said:

I'd say you're in a safe zone.  You may be able to push more, but I'm not the person to recommend it.  I think you would likely see great results in temp from a delid as I've never seen poor results from doing so on a 7th gen, but it is quite risky.  I've seen many examples of a 20°C drop with delid and liquid metal tim.
I'd also consider checking out the fan and pump speed on your AIO.  It seems like your temps are a bit high for the voltage you're stating.  It could also be that you're stating what you set it at and not what it is reported at during a stress test as the voltage does change with line level control.

Yes - I know about, and can see, the voltage variations. Easily .03V higher than setting for brief periods. As I understand it, that's under control of the motherboard. Maybe an attempt at stability.

 

I have the Kraken on 'performance' mode at all times, but we are talking NZXT Cam software - so who the fork knows what it's actually doing...

 

I will mess about with fan curves and the like. I agree that the temps were above what I expected, even as a rookie at this. Maybe I just got a crappy chip; maybe it is something in my settings. For know I am going to keep on tuning. It's as much fun for me as playing games!

 

And I knew all about the 7-gen chips and the issues with the cheap-ass tim. I cannot for the life of me understand why Intel doesn't just do a refresh of Skylake/Kaby Lake with proper bonding materials under the heat spreader, sell it for $10 more and market the crap out of it as a 'sooper-overclocker!'.

 

We would snap those babies up!

 

But who the hell knows how these guys think? I mean they spent several billion dollars researching and developing this tech and then crippled it at assembly to save a couple of bucks. The more I think about it, the more unfathomable it becomes.

[Cracklingice]

1 hour ago, Lwizzit said:

But who the hell knows how these guys think? I mean they spent several billion dollars researching and developing this tech and then crippled it at assembly to save a couple of bucks. The more I think about it, the more unfathomable it becomes.

Keep in mind too that overclock speeds on Sandybridge really aren't that far off from Skylake, but the Skylake parts were default clocked significantly higher.  hmmm

[greaves1111]

I bought the Kraken x52 cooler as well and installed it earlier this week. I also noticed really high temps stress testing at a fairly agressive OC. I didn't mess with voltages and used MSI's overclocking utility in the UEFI. Using Intel burntest, I went to 95c on a 10 minute stress test at 4.9ghz on a i7 6700k and it eventually crashed

 

Decided to bump it down to 4.7 and ran the same test for half an hour. Temps never went above 70 and it idles at 27c.

 

 

 

 

[Lwizzit]

On 4/28/2017 at 4:15 PM, greaves1111 said:

I bought the Kraken x52 cooler as well and installed it earlier this week. I also noticed really high temps stress testing at a fairly agressive OC. I didn't mess with voltages and used MSI's overclocking utility in the UEFI. Using Intel burntest, I went to 95c on a 10 minute stress test at 4.9ghz on a i7 6700k and it eventually crashed

 

Decided to bump it down to 4.7 and ran the same test for half an hour. Temps never went above 70 and it idles at 27c.

 

 

 

 

Vcore voltage is ?