Project i7 930

[aliasaid]

Meh, not really that great if you ask me. Plenty of D0's can achieve 4.2ghz on the same or less volts.

Well, nobody was expecting much higher...

This was just a test to see if it OC's like a 920 or not...

 

[Burn]

I know that, and Slay's results just reaffirm the idea that Intel crippled the 930 to make it's higher end CPU's look better.

 

[orihS ‪‪‪Shiro]

I know that, and Slay's results just reaffirm the idea that Intel crippled the 930 to make it's higher end CPU's look better.

Crippled?
Slay hasn't even gotten under the hood of this yet.
I mean, Slay's just played with the multiplier and a slight vcore bump.
Let SM get a better handle on this, I mean, it's only been one day.
I'm expecting good things. :]

 

[Burn]

Crippled?
Slay hasn't even gotten under the hood of this yet.
I mean, Slay's just played with the multiplier and a slight vcore bump.
Let SM get a better handle on this, I mean, it's only been one day.
I'm expecting good things. :]

Yes crippled, if you look at the majority of the 930's out there now not very many of them clock near as good as 90% of the 920's did.

Maybe I did jump the gun a bit though. I'm still hoping for good things.

 

[Slaymate]

Well I decided to try running OCCT (a prime number stress test) with the 930 at 4.2GHz (21x200) and it passed a hour run. But it still fails at LinX.

So I decided to try using the extra Multiplier the 930 has and see where it would pass LinX at. I started at (22x191) which equals 4.2GHz but it still failed. I started reducing the BLCK (or QPI) by 1 until it successfully passed LinX. This happened at 188, or what CPU-Z is showing as 4125.9GHz.

I've still got a couple of tricks to test out, I'm still shooting for 4.2GHz

 

[dirkpitt45]

The core's temps seem way far apart?

Unless I'm reading the wrong (just use coretemp personally, nice and simple ) but one's at 76 and the other's at 68?

 

[S0ULphIRE]

That's what I said too, slay are you sure everything's nice and flat there?

 

[Slaymate]

Everything is as smooth as baby skin and flat as glass. One way of controlling the speed and overclocking ability of a cpu is by controlling the air content of the thermal epoxy. A couple of well placed air bubbles and suddenly you have control of how much heat is being dissipated. Considering just how low core#3 is makes me think that the problem is below the heatspreader.

I went ahead and lowered a couple voltages and ran LinX at 4.125GHz again. I also had a couple of different monitoring programs running. EVGA-ELEET doesn't keep a Min. and Max. temp but you will notice it's reporting slightly different idle temps than HW Monitor or Real Temp and they (Real Temp and HW Monitor) are also different. Look at HW Monitor, it's reporting Core#3 as having a Max. Temp of 64C at 4.1GHz on Air. That's almost watercooling temps. If all the cores ran that cool I'd be shooting for 4.4 to 4.5GHz

 

[S0ULphIRE]

bump up the cpu vtt mebbe? Most 930 OC runs I've been looking at have it set a little higher what you've got.

 

[Slaymate]

Well today I decided to see what the 930 would do with HT turned off. At 4.2GHz it passed LinX with no problems, it even posted a new High Score for the LinX thread, 62.3938 GFlops. I also ran a quick 1M Super PI run and a 3DMark Vantage.

I then bumped it up to 4.3GHz and it also passed LinX, with another new High Score for the LinX thread, 63.5300 GFlops. And I made a Super Pi and Vantage run.

When I jumped up to 4.4GHz I hit another Stone Wall, I tried (22x200), (21x210) and (20x220). Tomorrow I'll start at 4.3GHz and try bumping the BLCK up by 1 and see how close to 4.4GHz it will get