i7 6700k: acceleration, tension, scalp
Recently became the proud owner of this gem and decided to try out its overclocking potential. Before that they went down on scalping (made vise, by the way, my first not successfully), got Zalman CNPS10x Performa plus, Asus Maximus VIII HERO and MX4 thermal paste.The stability of the dispersal checked a half-hour run LinX 0.6.5 m using 8GB memory. Percent I have got a VID 1.28 v, which is not the best option for overclocking.
4.6 Ghz took at a voltage of 1.30 v in the BIOS (under load in the system increased to 1.33 v), temperature up to 58C° at the peak.
4.7 Ghz is not taken until I put the voltage to 1.36 v (in the system to 1.39), was heated only up to 62C°.
The experiments stopped, returning to the 4.6/1.30.
The official specs of this processor shows the recommended voltage to 1.52 v, but I don't want to strain percent.
Share your experience of using 6700k 6600k and how did you do the VID, scalped than cool and what results the frequency and voltage is reached.
Sohrab
How were you able to get 4.6 at 1.3 V with VID 1.28 v I XS. I have 4.5 took 1.335 VID at 1.24 and 4.6 even at 1.4+ takes)
Listoman
all Protsiv different capacity , someone at the stock voltage of 4.5 takes , and have someone else on the 1.4... apparently successful instance populace or the person overpaid the money and bought the choicest percent , for example, I did so because in the background the total cost of this overpayment did not seem to overpayment...
WWQ
I bought the first b/u for 18K + 1K for the scalp. Relatively lucky to have him, it turns out.
And he showed himself in overclocking, the frequency/voltage?
Sohrab
well, I have a 6850 with starting a 1 volt level at 1.24 4.4 takes , but the test does not pass , at 1.29 passes all tests (30min) + allows a single core to lift up to 4.5 , and 1,331 is allows the core to lift up to 4.6 , thanks to turbo boost 3.0 + case at 1.33 to 3.8 , can be higher Keshi to drive but it makes no sense ,because even in benchmarks parrots adds , but in games even more so.
In short in single-threaded benchmarks at the level of 6700 4.7-4.8 GHz with this new Busta)
My 4790к ascalifornia takes 4.7 GHz at 1.275 V. 4.5 GHz takes at 1.2 V. However all this is not in the Linx test and OCCT test. The lynx do not see the point to test, because it is an inadequate heater.
PS: forgot to say that these tests drove vodanet Arctic Cooling Freezer 240, which not so long ago dead pump. :( Now sit temporarily on macho. Frequency of 4.7 GHz at 1.275 V. in games keeps within 60-75 gr. depending on the load. Today ordered another ready-dropsy, 3 days should arrive.
zy2: the rock of my 1.070 V.
6700K@4500Mhz - 1.245 V, OFF 0.005 V, 4 LeveL, AdaptiveMode, windows decided to VID max-1,265 V(HWInfo)(Not scalped, cooler Scythe Ninja3 Rev V, LinX task 40000 - 30min, the Temperature of the hottest core 77градусов in the room 27 degrees, in games max 55 degrees)Bolshii did not drive - don't see the point for everyday tasks is more than enough.
Pers4444 wrote:
In windows decided to VID max-1,265 V
Sohrab wrote:
Percent I have got a VID 1.28 v, which is not the best option for overclocking.
who look in Windows like that?) It is necessary in BIOS to look the same.
-SK.art - wrote:
who look in Windows like that?) It is necessary in BIOS to look the same.
well, in the dynamics in the BIOS will not see a load generally, the value is displayed in the BIOS , I like that I was doing in the load is +0.013 , and idle time varies between exposed in the BIOS and -0.091 him , is seen at the hardware level.
I'm here with these little stupid. Kaspalov and Silakov different values. Hastelow view is much lower than that of the good. If hastelow kind of 1.070-1.1 is considered very high, then Silakov they are in other limits.
At 6600k VIDы from about 1.1 to 1.2, have a 6700k from 1.2 to 1.3 V.
-SK.art-
My old 4690k also worked for 4.5 GHz, but LinX pierce did not pass. Sometimes got BSOD while playing, so yuzal it to 4.2 GHz.
The successful LinX walkthrough - an indicator of stable operation of the CPU, if mastered - mastered anything.
Sohrab wrote:
if mastered - mastered anything
the fact that links inadequate warm processor. The same Prime load the CPU is not worse, while not so warm much. My stone is Prime, SCR at a frequency of 4.7 GHz. And no badow in games.
WWQ
all Protsiv different capacity , someone at the stock voltage of 4.5 takes , and have someone else on the 1.4... apparently successful instance populace or the person overpaid the money and bought the choicest percent , for example, I did so because in the background the total cost of this overpayment did not seem to overpayment...
The choicest of the processor is expressed only in the form of a VID-a . The lower the VID the better the overclock. Great VID with the CPU not overclocking well...-SK.art-
The lynx do not see the point to test, because it is an inadequate heater.if I remember correctly , the SCR is a test this is the same Linpack only with avx instructions...
Prime need only to those who have the computer still work 24/7
Vinni-Pukh
I have identified 2 tests , first 5 minutes Aida (quick test on the blue screens) , if you pass , then 20min of the CFC , as this practice is enough.
And Prime is like the same of the CFC , only 5 degrees more warm (for me) , usually if the CFC passes the Prime and is now Prime is not in use.
WWQ
SCR is the same linpack and Prime95 uses FMA3
20 minutes to Prime it a little, at least an hour. In General, 10 hours at least.
Typically, the Prime need was 0.030 V more than linpack for stability
And for games will be enough only AID-s.
Vinni-Pukh wrote:
And for games will be enough only AID-s.
Thought so, too, in NFS 2015 and GTA 5 it doesn't work if the RAM is not very configured, although Hades is, but after osta , all in chocolate.
And these 0,030 V not sour on temperature, pressure , and iron so there is no need to rape, especially because it works, and to improve strained order in which the test error climbs IMHO nonsense)