Overclocking is all about overclocking v.6

Scalped your 4770k, took specu in these cases to Minsk. Did it before the new year, now waste, drove it. You can share the experience. Of course expect more, at least the reset temperature, the reality was sadder, as the master explained the crystal caught humpback and cap percent is not tight it is, plus the hot temper of my PCT VID 1.088 v, eventually replacing the thermal paste on the LM gave the win only 10C, two years of use the temperature has deteriorated by an average of 10C a GM returned the temperature was when buying percent. It turns out intelliscore G under the lid until it withered away completely, coped with its task, the XS but the results prove it, although on the same overah on the forum all say that a minimum of 17C GM fold, as you can see there are exceptions like my case and at least 10C. Stable overclocking in LinX 0.6.5 achieved only @4400 1.28 v, more trottling.

DozoRRR
as the master explained crystal got humped
If laughter prolongs life so you sing extended for at least a day.
@4400 1.28
Another day.

My dear friend, crystal of the processor is always flat as a mirror. In extreme cases, this may be the IHS is not perfectly flat, and it is unlikely for the inner surface. And then, what is this master that this cannot be correct? His skill is such cases?

And for goodness sakes, what did you do with hasholom he began to demand of 1.28 v for 4.4 GHz? He reported taking a frequency at 1.15 v or less.


You know, I had a good idea: every time when I see on the forum anything stupid, like these, I will copy them in the FAQ as question.
The country must know its heroes.

(Just kidding of course, Sergey, I will not allow this, but generally it would not be bad)

DartMaul
And what's so funny? You have to explain then. Describe in your own words how could, how much in the know. If you have the spices in it so it's funny, would give can advice.
For the sake of interest, I've seen people have i7 4790k how many volt requires for @4,4 Ghz?

Scalped BY_Pashka on overo, not like someone did not complain about his work. What he said revealed more than 400 percent.

DartMaul
And for goodness sakes, what did you do with hasholom he began to demand of 1.28 v for 4.4 GHz? He reported taking a frequency at 1.15 v or less.

no one and nothing hasfal don't have to... I Have my shitty 4670к takes 4.3 GHz at 1.28 V. VID processor 1.048.
There could be as lucky with the CPU. Not all successful. With 4690к/4790к usually doing better things..

DozoRRR
As You so dispersed? I 4770К work on 4.5 at all without any manual voltage increase, automatically up load to a maximum of 1.2 v With a mediocre cooler 4 heat pipe is heated up to a maximum of 67-70g S. And he worked me so perhaps a month, then I decided that enough with his violence and dropped to 4.2
Under the hood I have not the GM, and just a good silver thermal paste, I don't remember exactly what.

Belomorkanal
Like-like. No luck with percent. On lower voltage tests I BSOD 0x124, as read on overah necessary to raise the voltage, which ultimately helped to 1.28 v

DozoRRR
-Sergey Viktorovich-
First I want to specify the voltage that you call is what you set in BIOS as the voltage of the cores, or something that really served there? (set/given) It's actually a huge difference.

124ый the office is not always the lack of power on the cores. It may be something related to memory, cache, and IMC. 124ый can be treated with draping on Prots, but it is not always correct.
Pure Procopy office is 101.

no one and nothing hasfal should not
4.3 GHz at 1.28 V. VID processor 1.048.
Look, if the VID 1.05 percent, so the drain (at 3.7 at all, 3,9 on the ground), it needs to be stable at 1.05 v.
It turns out that only 400 MHz, you need 0.2 v more. (if you count from 3.9).
It is known that at room temperature Haswell takes 5GHz+. From this I conclude that the curve mult/vcore should start a break somewhere in the region of 4.5 to 4.6. Consequently 100MHz up to this point should require not more than 0,03 v.

In short, or you forgot to include the LLC, or in your room +50C, or you have a Performa cooler (hopefully thick enough to thin)
Of course there is a chance that it's just a very bad chip. But in this case no it is not something to disperse, and even to scalp. Generally on your place I would have just killed him and changed under warranty.
Repeat until something better comes up)

Belomorkanal
works on 4.5
And he worked me so perhaps a month, then I decided that enough with his violence and dropped to 4.2
Actually, I expect that there will be enough of this violence, and was raised to 5, but alas.
Spoileres acceleration

With overclocking

DartMaul
I don't want some nigger knocked from my hand food and grabbed my stuff. For this I'll be on 4.2 will stop for now.
Spoiler me too weak cooler to count on a bonus crackdown. And since I'm using a PC not to benci and Igor, I have enough of 4.2, because my video card is not the top, to the very demanding CPU Igor to pull more than 90-100ФПС. Now, if you decide to SLI sculpt (which is unlikely) two 980 or vidyahu change (most likely will) on what be GTX+100500ti or R9 490Хэ, then put the cooler on 6 tubes and I will be overclocking. Spitalnic the mother of the 8-phase, should be enough for any overclocking on air. Maybe even re-skin, I have quite a bit of sealant grease under the edge of the cover, to effortlessly it if you need to remove
PS I also think that they naprugu not watch it. Personally, I measured directly the most common multimeter pozaimstvovan (okay, to be honest, steal) from work (past practice and not past due).

Belomorkanal
I don't want some nigger knocked from my hand food and grabbed my stuff.
The analogy was exactly the opposite, but it does not matter.
Under the spoiler are all reasonable, without objection.

I also checked my mother a multimeter. Consistently multimeter shows 0.02 v more than the mother, and all voltages.

DartMaul
I hope SW SV will not be offtopic
After working with a multimeter should relate to the topic of overclocking the CPU, right?
If at all voltage the same 0.02 v is the error of the multimeter, Yes, I am Captain Obvious.
As far as I remember from the course of the verification officer (Oh davnooo was), a conventional multimeter accuracy class 0.5, thus acceptable (for professional use) the error should not exceed 0.06 v. To say nothing will, because, first of all, I have ITR and I personally do not work with KIP, and secondly, the course of the verification officer, verification officer and I never worked, I was just an engineer (a slave in the white helmet) with Assinica for me was the main thing that the device was not expired, was the mark of a verification officer with the previous year and the date of the next calibration.

DozoRRR
Or You have a bad copy or do something wrong, at the time, and 4770k and 4790k steadily kept 4.5, 4.7, 4.9 and last even when constantly working with no off-scale temperature. The same 5960 out of the box immediately without any problems with its standard 3.0, took 4.5 and never had any problems for 10 months, of which on / off the system was, God forbid, a couple of days to load (render / ue4) most of the time. All frequencies is achieved without the scalping, the only problem was in the first days of use 4790k, but not with him but with the motherboard, burned down when the CPU took 5.2, but this is nothing more than self-indulgence, by the way, it changed a month later under warranty, thanks to the ASUS-u)

Belomorkanal
After working with a multimeter should relate to the topic of overclocking the CPU, right?
It refers to the subject of measurement voltage)

If at all voltage the same 0.02 v is the error of the multimeter
But one of them is lying, the mother or multimeter, I think I have to learn not to be. But it does not matter.

Mr.Sirocco
by the way, it changed a month later under warranty, thanks to the ASUS-u)
Thanks to this when the site change, but after a month it's nasty dogs)

DartMaul
Whichever the event of a breakdown, then sent for testing :) was Not a problem of TC is still going system on 5960, and the mother was taken to the price increase of course and the new came after, in the end, the updated model for newer and twice as expensive)

Mr.Sirocco
Whichever the event of a breakdown
Despite what the warranty policy of the store.

Guys, Hello! I want to overclock my FX4100, cooler Titan

Tried to carry out the stress test to start for 10 minutes, the temperature levels off and does not grow further. But why such different graphics cores?

Guys help to overclock the FX 8300
The cooler scythe ninja 3 Rev. b
The temperature at the moment 34-36 degrees (during installation)
It is desirable to 4.5 GHz, but we'll see
what settings to change in the BIOS what to disable, etc.
With many grateful

ICYMEANSCOLD
For starters, what motherboard? She may not pull @4,5

DozoRRR
ga 970a ds3p

Fellow scholars, there is a problem with overclocking.
Don't want to disable analogsuicide technology hurts also I put the Vcore voltage on the mainland of ga 970a ud3p Normal and indicate dynamic voltage DVID + 0.0725 V. General to 1.325 V standard voltage for the CPU should give 1.4
Driven very rough, just a multiplier, put X19. In the end, the system is stable and runs well even on the test AIDAх64 and AMDOverdrive. The regular 140mm fan glacialtech cooler igloo 58 pwm keeps the CPU within a 52-53S°
However, in Fallout 4 at the frequency of 3.8 happens to BDOS with a red screen, with no error messages not displaying, just a red screen, the crackling of the speakers and a separate computer restart in 5 seconds.
What would this red screen mean? Whether a memory conflict (I have 4 modules whales for 2 straps of different manufacturers are) or something is overheating (going to overheat all the sensors far away) and even cooler food not heated above 50°C (I have it even the CPU fan blows gently)
Blue screens had not been caught, even red-catching, but they were writing with codes, though, could understand what stopped the system. And then, without inscriptions
PS
Also don't understand why the hell we had to do the configuration of the two memory channels separate? I used to have these straps on 1600 worked, now the motherboard reads from the profile data by default (old do not know how and put the timings at most on all 4 straps) and one data set for 1600 no=( In the end, auto-tuning leads to the fact that more of the old memory just is not determined, and setting the 4 timings is clearly not sufficient for its stable operation. (it turns out that additional timings of one of the pairs of laths exhibited at 1333 and differ from those in the second pair at a frequency of 1600)

Iron really dead?