Ivy Bridge Results and Discussion thread

[Rvenger]

Also, for some reason, sometimes my voltage starts fluctuating and goes anywhere from 0.9 to around 1.12~1.14; When running Prime95 it stays still on 1.096, any idea why this happens?

vdroop

Basically its using what it needs. So if prime requires 1.096 depending on the load, it will use it. If you want to see max voltages I notice if you run the Winrar benchmark test, it likes to spike the voltages up high.

 

[Avalon]

Seems too low for 4.1Ghz

Hmm really? I was under the impression that it was too high. I made some tweaks to PLL and VCCSA voltages and managed to get it stable at slightly less volts.

 

[Don Karnage]

Hmm really? I was under the impression that it was too high. I made some tweaks to PLL and VCCSA voltages and managed to get it stable at slightly less volts.

How far did you pull the pll down?

 

[IntelUser2000]

Have you guys seen this link? PCWatch did a test with new thermal interface material, but rather than testing it bare die, they put the heatspreader back on:

http://pc.watch.impress.co.jp/docs/column/sebuncha/20120511_532119.html

Temps: http://pc.watch.impress.co.jp/img/pcw/docs/532/119/html/graph1.gif.html

8-10C reduction at stock and 15-20C reduction with overclock.

 

[QuantumPion]

I thought my system was stable with 4.5 Ghz at 1.21 V but it was still crashing on overnight OCCT runs. I had to bump my offset up to 0.06 (~1.23-1.24 V) to stop the crashing. Temps are high but still acceptable to me, ~60-65 for normal programs, ~70-75 for stress tests. I'm still unable to overclock my ram at all (4x4 GB Samsung 30 nm 1.35 V).

 

[dma0991]

Have you guys seen this link? PCWatch did a test with new thermal interface material, but rather than testing it bare die, they put the heatspreader back on:

Intel's gonna use solder with Haswell and all of a sudden the temp issue is solved with 22nm FinFET. :whiste:

 

[jp0ll]

So I was able to run 20 passes of LinX(a little over 3 hours) with all memory selected with no problems at 4.2Ghz and 1.12 vcore. Tried running Prime95 blend overnight and I woke up to find that the system had restarted about 5 hours in. Should I try bumping the vcore and trying again?

 

[dma0991]

So I was able to run 20 passes of LinX(a little over 3 hours) with all memory selected with no problems at 4.2Ghz and 1.12 vcore. Tried running Prime95 blend overnight and I woke up to find that the system had restarted about 5 hours in. Should I try bumping the vcore and trying again?

Yes, you'll have to do it a couple of times till you achieve a voltage that you think is stable. 1.12V is a little bit low for my chip and if not mistaken I did 4.2GHz @ 1.18V but it varies from one CPU to another.

 

[exar333]

Have you guys seen this link? PCWatch did a test with new thermal interface material, but rather than testing it bare die, they put the heatspreader back on:

http://pc.watch.impress.co.jp/docs/column/sebuncha/20120511_532119.html

Temps: http://pc.watch.impress.co.jp/img/pcw/docs/532/119/html/graph1.gif.html

8-10C reduction at stock and 15-20C reduction with overclock.

Very interesting...

 

[jp0ll]

Yes, you'll have to do it a couple of times till you achieve a voltage that you think is stable. 1.12V is a little bit low for my chip and if not mistaken I did 4.2GHz @ 1.18V but it varies from one CPU to another.

Sounds good. I'll bump up the voltage tonight and see how it goes overnight.

 

[Don Karnage]

Have you guys seen this link? PCWatch did a test with new thermal interface material, but rather than testing it bare die, they put the heatspreader back on:

http://pc.watch.impress.co.jp/docs/column/sebuncha/20120511_532119.html

Temps: http://pc.watch.impress.co.jp/img/pcw/docs/532/119/html/graph1.gif.html

8-10C reduction at stock and 15-20C reduction with overclock.

While nice the mhz wall is still around 4500. After that huge voltage jumps are needed for stability.

 

[Avalon]

How far did you pull the pll down?

My Asrock Extreme4 was set to auto and it was reporting as 1.856V, I dropped it to 1.79V.

 

[Ferzerp]

While nice the mhz wall is still around 4500. After that huge voltage jumps are needed for stability.

They claim 4.7 with 1.25V in that link.

 

[Don Karnage]

My Asrock Extreme4 was set to auto and it was reporting as 1.856V, I dropped it to 1.79V.

I run mine at 1.25v. Try dropping it down more.

 

[Rvenger]

I run mine at 1.25v. Try dropping it down more.

It amazes me that you are not crashing with that low of a PLL.

 

[baik]

I recently got my 3570k w/ the asrock extreme4 setup going and I'm getting 44x with my vcore at 1.152v with temps maxing out at about 70c running a 212+ on it

tried to go 45x without upping the voltage but things seem to go south pretty fast...oh well

 

[dma0991]

I recently got my 3570k w/ the asrock extreme4 setup going and I'm getting 44x with my vcore at 1.152v with temps maxing out at about 70c running a 212+ on it

tried to go 45x without upping the voltage but things seem to go south pretty fast...oh well

That sounds like a pretty good chip, given the fact that it needs so little voltages for 4.4GHz. How long did you run stable Prime95 at that settings?

 

[Xeris]

So I just got around to messing around with my 3770k and it seems like if I try to go to 4.7GHz, I get crashes until I bump up my voltage to over 1.3V. At that point the temps hit about 90C and I just don't feel comfortable with my CPU running that hot :\ I just decided to go down to 4.6GHz at around 1.27V and keep it at that because my temps are lower than 85C on load for around 30minutes. I'm going to run Prime overnight and see what happens when I wake up. If it is stable then, I will run a few other tests and then settle

 

[SHAQ]

So I just got around to messing around with my 3770k and it seems like if I try to go to 4.7GHz, I get crashes until I bump up my voltage to over 1.3V. At that point the temps hit about 90C and I just don't feel comfortable with my CPU running that hot :\ I just decided to go down to 4.6GHz at around 1.27V and keep it at that because my temps are lower than 85C on load for around 30minutes. I'm going to run Prime overnight and see what happens when I wake up. If it is stable then, I will run a few other tests and then settle

Remember to get the marshmallows before you set it up for the night!

 

[Tempered81]

Intel's gonna use solder with Haswell and all of a sudden the temp issue is solved with 22nm FinFET. :whiste:

I think they'll use it on IBE as well. 6-Core is going to be hotter anyway, so that should be able to keep the overclocking limits on par with IB

 

[Chaotic42]

I got my 3770K in yesterday morning. I'm starting the build on it now. Maybe this will break my overclocking curse? It's batch 3213B589 I guess that's week 13? It's from Costa Rica, they could have at least put some of that awesome coffee in the box....

Edit: Yay, this is the only relevant Google result for that batch number. I'm a trailblazer.

 

[dma0991]

I got my 3770K in yesterday morning. I'm starting the build on it now. Maybe this will break my overclocking curse? It's batch 3213B589 I guess that's week 13? It's from Costa Rica, they could have at least put some of that awesome coffee in the box....

I have some bad news for you.

 

[Chaotic42]

I have some bad news for you.

F..... okay . Stupid curse.

 

[darkxshade]

With the 212+ on my 3570K with Z77 Asrock Extreme4:

Got 4.5 easily with 1.2v using .01 offset & LLC3
P95 Stable with temps in mid 70s, IBT in mid 80s(core 2 max @ 87)


Like others said, that's where the wall is. I had to tweak through multiple BSODs to get 4.6 which required ~1.23v with .04 offset & LLC3. IBT temps hit about 94 max on core 2.

No point trying for 4.7, will probably pull back down to 4.5 since the extra juice to get 4.6 is just not worth it.

 

[Xeris]

Turns out my overnight test was a fail It crashed around 3 hours in. Bumped up the voltages a tad and going to leave it for the day. Currently at 4.6GHz @ 1.28V. I'll let you guys know what happens after work today