3770k @ 4.3?

[aaksheytalwar]

Hi

I will be buying a 3770k soon, hopefully I want to run at 4.3, (running my SB at 4.3 too). Because after 4.3 the volts/temps start increasing quite a bit without any noticeable performance advantage. If I get a really good chip I may do 4.4 or so, but I don't wanna go below 4.3 as that is what I currently am with my SB.

What is the max volts any 3770k would need to do 4.3 100% stable? Is 1.15 volts realistic or optimistic?

I hope I can stay below 1.2V. Is that sureshot?

 

[exar333]

I would think most 3770Ks will do 4.3 at 1.2v or less. Temps should be manageable at that freq as well with your H100. You probably could kick it to 4.5, assuming you get an average OCer or better.

 

[MPiland]

Hi

I will be buying a 3770k soon, hopefully I want to run at 4.3, (running my SB at 4.3 too). Because after 4.3 the volts/temps start increasing quite a bit without any noticeable performance advantage. If I get a really good chip I may do 4.4 or so, but I don't wanna go below 4.3 as that is what I currently am with my SB.

What is the max volts any 3770k would need to do 4.3 100% stable? Is 1.15 volts realistic or optimistic?

I hope I can stay below 1.2V. Is that sureshot?

Well, right now I'm at 4.5 using 1.21v, so I'd assume you could run lower than that to stay at 4.3. I've heard of some people getting to 4.2 at stock voltage if that's any indication.

 

[LagunaX]

I would think so, we've been seeing 4.5ghz stable at 1.2-1.24v on 3770k's.

 

[Stochastic]

This is from the Anandtech article on Ivy Bridge overclocking. Ian Cutress was able to hit 4.5 GHz at 1.1 volts. I'm not sure if these results are realistic since he was using liquid cooling. Would it be unreasonable to expect to reach 4.3 GHz@1.05 volts? This is what I currently plan to do with my 3570K when I begin the build. I might increase the voltage to 1.1 or settle for 4.2 GHz just to be on the safe side, though.

 

[MPiland]

This is from the Anandtech article on Ivy Bridge overclocking. Ian Cutress was able to hit 4.5 GHz at 1.1 volts. I'm not sure if these results are realistic since he was using liquid cooling. Would it be unreasonable to expect to reach 4.3 GHz@1.05 volts? This is what I currently plan to do with my 3570K when I begin the build. I might increase the voltage to 1.1 or settle for 4.2 GHz just to be on the safe side, though.

I think his was an ES chip or he got lucky. Most people have been needing 1.2-1.24 for a 4.5 OC while he's showing much less than that. You could probably do 4.3 at SLIGHTLY above stock.

 

[aaksheytalwar]

Can i get full load temps below 65c at 4.5?

 

[MPiland]

Can i get full load temps below 65c at 4.5?

I guess that depends on your cooling really. I have the H100 and at full load, I'm in the mid 70s which seems about par to most chips OC'ed that high.

 

[aaksheytalwar]

Prime95 or IBT?

 

[dma0991]

Can i get full load temps below 65c at 4.5?

Not possible unless you have ambient temps that are lower than mine, 30C or ~26C at night. Considering that I'm running a custom loop, the H100 should not be able to achieve 4.5GHz @ 65C.

I'm running 100% load at 4.5GHz @ 1.28V and the temps are near 70C if it is on a hot day. 60-65C on a cold night. Since voltage requirements increase with temp, if I were to run it 100% load and it exceeds 70C around the 75C region, it will crash the program. Anything below 70C is rock solid.

 

[aaksheytalwar]

Isn't 1.28v too high for 4.5? I am thinking sub 1.25v

 

[MPiland]

Isn't 1.28v too high for 4.5? I am thinking sub 1.25v

Not really. It all depends on the chip. Each chip is different. I've seen people get 4.5 at 1.18v, but it took me and a lot of others ~1.22v. It's just the luck of the draw. And my temps are on Prime95 on the long run. I'm going to run a blend test tonight after I do some more tweaking.

 

[dma0991]

Isn't 1.28v too high for 4.5? I am thinking sub 1.25v

Not every chip that rolls out of the factory are made equal and from what I've seen so far, Core i7 3770K needs less voltage than Core i5 3570K at the same clockspeed.

As far as 1.28V goes, it doesn't bother me as it is still below the 1.3V mark and it is from then onwards where my voltage requirement spikes if I want to attain a higher clockspeed. Durability isn't an issue for me as I run stock most of the time. I know exactly when I need to run it at 100% and a quick restart and load OC profile is not a hassle to me.

 

[aaksheytalwar]

Can 4.3 be managed sub 1.2v in that case with sub 60c?

 

[dma0991]

Can 4.3 be managed sub 1.2v in that case with sub 60c?

I'm not quite sure about 4.3GHz but I've tested 4.2GHz with auto voltage which is just a little bit higher than original stock but but still below 1.2V. You could probably go with less volts with the Core i7 3770K.

Sub 60C at that settings is difficult to tell as it varies with a lot of factors but I wouldn't be worried as 60C is quite low. I wouldn't be worried if it did 70C+-5C and it only needs low volts, stability is of the highest priority.

 

[MPiland]

Can 4.3 be managed sub 1.2v in that case with sub 60c?

Again, it's hard to answer since each chip is different. You could probably get under 1.2v, but the temps I'm not so sure. That really depends on the ambient temp, case airflow, cooler solution, etc.

 

[Rvenger]

IMO hold onto your 2600k. I found out that my 2700k was an above average chip via PM this morning.

 

[aaksheytalwar]

Will be getting a 3770k regardless

 

[JAG87]

Can 4.3 be managed sub 1.2v in that case with sub 60c?

I am running 4.3 at essentially stock voltage (+0.005V offset), which is ~1.080V under Linx and P95 load, and just over 1.1V under normal load. It's 8+ hour Linx and P95 stable. 4.4 was not stable without increasing the offset and it's just not worth it to me.

Temperature wise I can't give you any reference since I'm semi-passive and my temps are around 75-80C under Linx (of course much more reasonable under normal load), but even with an H100 I don't think you'll be seeing less than 60C in Linx, I'd guess probably around 60-65. Under normal load, probably 50-55C.

 

[chloros]

running 4.4 now at voltage (-0.005V offset) which caps the voltage at 1.84v and PLL set to 1.6v. Once i go to 4.5 things heat up and it requires 1.2v.

Temps at 4.4 will max around 79C on 1 or 2 cores. Idle in the high 20's

I would imagine 4.3 would be a breeze.

 

[MPiland]

running 4.4 now at voltage (-0.005V offset) which caps the voltage at 1.84v and PLL set to 1.6v. Once i go to 4.5 things heat up and it requires 1.2v.

Temps at 4.4 will max around 79C on 1 or 2 cores. Idle in the high 20's

I would imagine 4.3 would be a breeze.

I really hope you meant 1.084v

 

[aaksheytalwar]

With my current setup I get 2600k 4.3 1.27v full load around late 50s C max. If I run for hours then perhaps Early 60s with occt

 

[chloros]

I really hope you meant 1.084v

I was just going off what cpuid was reporting when running full load. Here is a pic.

 

[MPiland]

I was just going off what cpuid was reporting when running full load. Here is a pic.

No, I know. I was making a joke because you mistakingly (I assume) said it capped your voltage at 1.84v. That's a ridiculous amount.

 

[IntelEnthusiast]

On my Intel® Core™ i5-3570K I got 4.2GHz at 1.2v. So I think it is realistic to think you might be able to get to 4.3GHz at 1.2v. Of course not all chips will perform the same but mine seems to be fairly average.