[LoneWolf121188]

I'm at 241x11 and 1.664 vcore (from CPUZ, BIOS is 1.6375) and I'll prime overnight...I can go up to 1.65 in the BIOS, but that means it will really be something like 1.68...how much is too much? If you temps are low, then you should be OK, right? 32C idle, currently (though how accurate that is, idk...my hardware temp sensor is reading 27.3, but thats taped to the side of my XP-120...probably more like 36-38C).

I'm trying to hit 2.7GHz (246x11) with a good amount of prime stability...I cant pass prime for more than 3 or 4 min when I try to get close to 2.7 though.

[TimoneX]

1.68v shuold be alright as long as your temps are under control.

[Binman_UK]

I'm trying to hit 2700 myself with my S747 Newc 3400+. I can get it to 2650, but i get errors in prime when i set it any higher.
The temperatures go above 60 degrees if i set the fsb / voltage any higher than i have it already though .

Let me know how you get on, coz it seems likely that we're hitting a similar kind of o/cing wall!

Binny

[LoneWolf121188]

I ran prime overnight and its still going at 241x11 with 1.664 vcore (1.6375 BIOS).

[LoneWolf121188]

well, 23 hours of prime at 241x11 and 1.664...should I go higher? Prime quit at 242x11 after a couple of hours...should I push the vcore to the top?

[TimoneX]

Sure go for it bro, as long as it's staying kewl. I'd be hesitant to run it at 1.7v 24/7, but I don't think testing it at that vcore will kill it if you keep it cool.

Nice OC btw.

[LoneWolf121188]

I hope you're right, man...1.680 vcore and 244x11 (Its actually 243 in the bios, but for some reason, CPU-Z is reading 244). Looks like I'm right around 33-34C idle. Here we go again...

[/EDIT] BSOD at 244x11 and 1.68. at 242x11, my temps hit 56C after about 5 min with S&M, so I guess its back down to 241x11 and 1.664 [/EDIT]

[TimoneX]

Yeah 56c is getting up there. Giving that XP120 quite a workout eh?

[LoneWolf121188]

yeah, no kidding. I need a Delta 3 blade, or even better, H20.

[TimoneX]

H2o

[kukyfrope]

Finally setup the WC huh TimoneX, temps are amazing wow! 20C! I'll keep following your DFI Lan party thread

[LoneWolf121188]

52C Load, and thats while running S&M. It was Prime stable for 23 hours here (241x11), so I think thats where I'll leave it.

[EspElement]

well i dont htink high vcore is bad as long as its cooled.. i ran my 2400+M at 2.0volts.. on air.. temps were 48C load.. i try to stay under 50 as much as possible... cept when i run S&M or something

on 1.525v 250*10 i get a 43C load on my stock heatsinc.. i ordered m y xp-120 and the delta 220cfm fan so im pretty damn set.. haha

btw i looked into watercoolings... like thermalrights.. and they temps of a slk 900U or wateverthe one i used for the barton cores.. the water temps were higher.. just that it was a hella lot quiter.. considering i used a 92mm tornado :-D


also i dont know about you guys... but for my DFI lanparty when i run sandra for my temp tests.. it says my chipset is 35C (hinting thats my cpus temp) and saying my cpu temp is like 30C.. when i run prime.. my chipset temp goes up.. so im almost curtain that the chipset temp is my cpu.. for sure.. and looking that he has a DFI his things could be mixed up 2... saying his cpu is 36C... why would ur chipset be at 36C reguardless!?

[LoneWolf121188]

^^ the NF4 chip gets hot FAST. That sounds about right to me, I think mines 32C.

2 volts? on air??? wow, I'm impressed.

I really need that delta...I think my fans somewhere around 90 cfm...of cource, mine does have blue LEDs, which is awesome.

[TimoneX]

2v on a barton is OK if it's cooled, I too ran close to that to achieve 2.72Ghz, but 2v on a winchester is asking for trouble, I don't think the chip would last long at that voltage.

My temps are most definately not reversed. The chipset gets hot, look at the bios the default settings for the chipset fan don't even have it going to 100% speed till 50c as I recall. I believe 36c is actually a pretty good nF4 temp for running 295x9 SLI, but I've never messed with the chipset fan controls either and as you can see it was doing 3668rpm when I snapped that pic. If you check around you'll find alot of guys reporting chipset temps over 40c at idle.

If you had a 900U equal or surpass water temps you either had air in the lines or a poor water cooling setup.

[GunnerMan]

You really make me want an XP-120 now lol. Great clocks and GL with 2.7!

[EspElement]

i got my xp-120 !! came in 2 days... im up and idle its 31C on load its 37C

same clock and volts as my sig :-)

thats without the as being set.. thats right when i fired it up! haha

[EspElement]

the water setup i was looking like i said was the thermaltake one.. i mean come on.. it cant be that bad is it?

i didnt actually own it.. i looked at reviews and **** at tomshardware.com

[TimoneX]

Thermaltake makes garbage. I'd bet $100 my $100 custom built water cooling setup would eat it alive.

[hitechjb1]

Increasing voltage is for getting higher frequency (economic return), not for the sake of higher voltage.

Increasing voltage will generally make the chip to run faster as the switching current for each transistor is higher. But increasing voltage also increases power of the chip, hence higher average temperature, higher temperature gradient within certain area of the chip. Higher temperature will then slow down the chip.

The delicate balance between voltage, frequency and temperature at maximal overclocking will NATURALLY determine the highest voltage and temperature for a given cooling.

Every chip is slightly different, it is hardly to put a priori a max voltage number for each chip.

The rule of thumb I use is for each step of voltage increase, if the increase in frequency is less than 0.4% for each degree C increase in temperature, maximal overclocking is near and not much more frequency can be practically obtained even voltage is pushed further.

E.g. at 2500 MHz
0.4% of 2500 MHz = 10 MHz
So if voltage is increased by a step, less than 10 MHz is obtained for 1 C rise in CPU temperature under load, then I think maximal overclocking can be considered achieved.


Voltage, temperature and frequency: the basic variables of overclocking (page 20)