Look this, i7-3770K my almost on default voltage (1.165V) run LinX AVX Edition on 4.5GHz... That was testing, not RealBench and AIDA64... Now i7-4770K and 4790K with some of serious tests 90C on default clock... You see here 4.5GHz less than 75C..LinX AVX..last linpack for that time.
http://imgur.com/2boU3ZT
Still is with me and many people want him... I'm sorry because I don't have Maximus V Extreme or V Gene I watercooling.
I would set 1.400V and probably get 5.0GHz stable. I push him to 4.8GHz over night in Prime95 and I couldn't more because temps...
And than pass 92-93C but didn't show sign of throtlling at all.
After that I stop to push him further and keep him 3.5 years on 4.5GHz on 1.200V and I nicely prevent degradation...On 1.800V he could run LinX, OCCT, Prime95 days... But I follow Intel steps to keep always little more just in case.
Mostly with i7-3770K first higher ammount of voltage need on 4.6-4.8GHz, usually is that first big difference compare to stock voltage.
Sometimes For 4.6 I was need to move on 1.225V... than to reach stabiluty on 4.8GHz need again 100mV and probably for 5.0GHz need agaim extra voltages 100-150mV and watercooling. I will sell him soon becuse I will need money for GPUa and other things...
I can't find cheap Maximus V Gene, I would build small backup RIG with him, I have memory DDR3, only board is
E-ATX and I would like something smaller. That was best Intel mainstream socket, 1155
Sandy and Ivy. One cooler, other better features and IMC.
Overclocking is not problem when you have nice hardware. But I never had such habbit to torture hardware.
I think about them similar like a alive being, I can't even look video clips when someone destroy Apple Phones on bend testing and similar things. I close my eyes. Now situation is much different because people could buy good samples from Silicon Lottery if they wish.
Personally I would sell example golden sample of i7-6700K for i7-5930K normal sample or similar situation...I never felt satisfaction to someone look some insane number in my signature. With this CPU was on 4.5GHz, and small number of people knew that he work almost on stock voltage on 4.5GHz while they hold 4.7 on 1.4V.
He had luck because I was owner because many people will push him so far until degradation process start.
http://imgur.com/2boU3ZT
Still is with me and many people want him... I'm sorry because I don't have Maximus V Extreme or V Gene I watercooling.
I would set 1.400V and probably get 5.0GHz stable. I push him to 4.8GHz over night in Prime95 and I couldn't more because temps...
And than pass 92-93C but didn't show sign of throtlling at all.
After that I stop to push him further and keep him 3.5 years on 4.5GHz on 1.200V and I nicely prevent degradation...On 1.800V he could run LinX, OCCT, Prime95 days... But I follow Intel steps to keep always little more just in case.
Mostly with i7-3770K first higher ammount of voltage need on 4.6-4.8GHz, usually is that first big difference compare to stock voltage.
Sometimes For 4.6 I was need to move on 1.225V... than to reach stabiluty on 4.8GHz need again 100mV and probably for 5.0GHz need agaim extra voltages 100-150mV and watercooling. I will sell him soon becuse I will need money for GPUa and other things...
I can't find cheap Maximus V Gene, I would build small backup RIG with him, I have memory DDR3, only board is
E-ATX and I would like something smaller. That was best Intel mainstream socket, 1155
Sandy and Ivy. One cooler, other better features and IMC.
Overclocking is not problem when you have nice hardware. But I never had such habbit to torture hardware.
I think about them similar like a alive being, I can't even look video clips when someone destroy Apple Phones on bend testing and similar things. I close my eyes. Now situation is much different because people could buy good samples from Silicon Lottery if they wish.
Personally I would sell example golden sample of i7-6700K for i7-5930K normal sample or similar situation...I never felt satisfaction to someone look some insane number in my signature. With this CPU was on 4.5GHz, and small number of people knew that he work almost on stock voltage on 4.5GHz while they hold 4.7 on 1.4V.
He had luck because I was owner because many people will push him so far until degradation process start.
People talk that Sandy Bridge is better overclocker and they mostly think on i5-2500K.
I7-2600K and i7-2700K mostly go on 4.6-4.7-4.8GHz... And they usually compare lower temperature from Sandy and their flux solder but forgot to say that Sandy was more sensitive on temps on over 80C. From other side Ivy worker normal and on 90 without throttling...
I had some ex screen shoots with long Prime tests on 4.8GHz...
I even think that I forgot to lock voltage offset and than even board increase voltage little for 0.015-0.020mV.
But I didn't want to torture tests any more... Same as simmerhead cooler was of H100. First 240mm AIO.
I remember in AIDA64 x48 multi on Ivy Bridge i7-3770K was exactly 37% OC.
But I didn't have money to invest in rea watercooling and than I was hold him on 4.5GHz and that was ideal clock for Ivy Bridge.
Everything over is torture.
http://imgur.com/ENgqieQ
http://imgur.com/NnQqawE
Later when I listen people avoid to talk that Sandy Bridge is better OC than Ivy.
Maybe average 100MHz and average 10C lower temps. That was Sandy advantage i7 vs i7.
Lot of Sandy owners present Ivy as completely disaster and many enthusiasts decide to wait i7-4770K immediately after first test of i7-3770K. I knew that situation with him only could be worse not better and than arrive hottest Intel ever. 80C in Prime95 AVX on stock without Turbo. Decision to not wait him was one of better in my life.
Anyways was slower than my i7-3770K on 4.5GHz and couldn't resist because higher temps in AVX stress tests to beat him on 4.8GHz. OK 4770K could offer same results on 4.5-4.6GHz but no one didn't crazy to start Prime95
AVX, LinX or OCCT with him on 4.5-4.6GHz.
I7-2600K and i7-2700K mostly go on 4.6-4.7-4.8GHz... And they usually compare lower temperature from Sandy and their flux solder but forgot to say that Sandy was more sensitive on temps on over 80C. From other side Ivy worker normal and on 90 without throttling...
I had some ex screen shoots with long Prime tests on 4.8GHz...
I even think that I forgot to lock voltage offset and than even board increase voltage little for 0.015-0.020mV.
But I didn't want to torture tests any more... Same as simmerhead cooler was of H100. First 240mm AIO.
I remember in AIDA64 x48 multi on Ivy Bridge i7-3770K was exactly 37% OC.
But I didn't have money to invest in rea watercooling and than I was hold him on 4.5GHz and that was ideal clock for Ivy Bridge.
Everything over is torture.
http://imgur.com/ENgqieQ
http://imgur.com/NnQqawE
Later when I listen people avoid to talk that Sandy Bridge is better OC than Ivy.
Maybe average 100MHz and average 10C lower temps. That was Sandy advantage i7 vs i7.
Lot of Sandy owners present Ivy as completely disaster and many enthusiasts decide to wait i7-4770K immediately after first test of i7-3770K. I knew that situation with him only could be worse not better and than arrive hottest Intel ever. 80C in Prime95 AVX on stock without Turbo. Decision to not wait him was one of better in my life.
Anyways was slower than my i7-3770K on 4.5GHz and couldn't resist because higher temps in AVX stress tests to beat him on 4.8GHz. OK 4770K could offer same results on 4.5-4.6GHz but no one didn't crazy to start Prime95
AVX, LinX or OCCT with him on 4.5-4.6GHz.
