View Full Version : Any Intel QX users on VapoChill or Prometeia?
CrashOv3r1De
11-15-2007, 04:21 PM
Hey guys I'm interested in anyone running thir day-to-day quad cores on a vapo rig? Could you please post your configurations and results
Thanks
Circaflex
11-15-2007, 04:56 PM
http://forum.asetek.com/showpost.php?p=80160&postcount=12
http://www.xtremesystems.org/forums/showthread.php?t=130922&highlight=quad
weescott
11-15-2007, 05:06 PM
My god! Do not post about Pazza quad core phase!
If you want more info PM me.
I wouldn't boughter buy a new unmodded LS, they don't hold quadload proper, maybe ~0 -10 range, I've test both mach and LS and neither of them holds below 0 @230w.
I have one modded LS and it does -26 @ 230w which is pretty much more then a 45nm quad produce
I have also mod mach, it holds -31 @ 230w, -37-38 @ qx9650 @ 461*9 1,54vcore.
I havn't got the time or possibility to run the LS on quadcore yet but as far as I can see it will hold it below -30.
well as far as i can se, i wouldn't bougther buy a massproduced LS or mach, better go for custom since it need to be modded anyway.
regards
Tim
jinu117
11-15-2007, 05:21 PM
In all honesty, it depends on what you run at. If you don't plan to run truely high voltage I think Vapochill LS can handle in range of 1.4-1.5v on Quad G0 fine and I am sure overclock on it will be much better than air or water :)
If you start pushing voltage beyond that, it's wild guess. I've seen vapochill LS holding anywhere from 175-200w tops before it runs out of steam (mind you full load as in 4 quad prime running for hours). The evaporator is excellent (copy of Baker evap, I figured you would remember ;P) it's just they used last wee bit of compressor which is tad small sized for current CPU at high voltage.
If you can grab one used cheap, I can't see any reason not trying it out though ;)
a lot of the time what happens is that either the compressor overheats from working too hard, or due to the charge and cap tube length, floodback occurs, so I wouldn't recommend buying a new vapo for a quadcore. I would either look into buying a unit from a builder, or buying a used vapo on the cheap and then paying someone to fix it to your needs
here's a great place to buy your parts from
http://www.under-the-ice.com/
[XC] gomeler
11-15-2007, 08:04 PM
floodback shouldn't ever be an issue on a stock tuned LS/Mach II, now once it's modded floodback will be an issue but that's only afterwards. I agree with Jinu on the capabilities of the Vapo and Mach II, they'll hold up rather nicely for dual-cores and even hold a quad decently but typical phase-change users want more performance than the stock units will absorb. Either find them used for sub-$400 and mod them or contract out a custom build though. If buying used go with a VapoLS over a Mach II, the case and evap in the LS > case and evap in the Mach II. Compressor is negligible, NF5.5CLX vs NF9FX when compared to cramming a nice DSH + condenser into the Mach II case.
terente0081
11-16-2007, 02:17 PM
I just hope I'm not speaking too soon.... I have my stock Vapochill LS cooling my QX6700 ES since July.
CPU voltage is set at 1.4725 for 3.6GHz day-to-day. It can do higher but I'm not comfortable, knowing my LS might not like the extra stress.
When idling, the evap temperature fluctuates between -48 and -51 (no speedstep etc).
Gaming and prime testing, highest I've seen was -36.
EDIT: BTW, my LS may also run hotter because I have it mounted in my LL V2000 case, along with the HDDs. Picture: http://picasaweb.google.co.uk/terente.xp/Calc/photo#5112847675752993634
little optimistic temps there terente0081,
I've been dealing with ls and they don't get below -42 -43*c (measured with K-probe) without load, that means I don't beleave your -36 either is 100% true either
maybe you've read those temps on LS display? then it would explain why they are so optimistic/low.
[edit]
Language
jinu117
11-16-2007, 03:17 PM
tim... he is reading it from vapochill controller. -36c is reasonable (about 3 degree off typically at that point, -48 -51c would be about 4-6 degree off at that point). Yup, since I do mod a lot, I have k-type next to it when I mod and notices this. Beyond that point, it could be as much as 10-20c off -_-;
So, 33c with about 135-150w load sounds about right to me?
Just don't get it warmer than that for prolonged time terente. That's where I believe you are starting to reach not so friendly zone for vapochill LS (where compressor shell gets up to 80c region for about 3-4 hours of use, I believe your curren situation if having good flow should be around 75c compressor shell).
Thanks for sharing terente :)
terente0081
11-16-2007, 07:02 PM
tim, that's the reading from the vapochill display. I don't know how to check the actual core temp since coretemp doesn't know how to display negative temps.
Is there any other tool I could use?
Jinu, that's the maximum I use for day-to-day apps and games.
I was wondering... would I be able to hold 4GHz on a QX9650 for daily use? Knowing it's 45nm and perhaps not as hot as kentsfield.
Clock to clock, how much cooler do these new yorkfields work compared to kentsfields?
jinu117
11-16-2007, 07:07 PM
We won't know till someone gets hand on one of those puppies and try it out.
However, I don't think 4ghz would even be a sweat considering I was able to get that speed with my 200w tuned (albeit colder running than stock Vapochill LS) quad prime stable on 4.2ghz range with Q6600 G0....? Newer chip should have really more headroom than 4ghz to be honest. I was thinking more in 4.5-5ghz range with newer CPU with decent SS unit... but won't know till I can try one.
Terente, unless your room typically stays above 27c I think your unit is probably what I call hold 175w-180w range unit which I saw most commonly. I think you are on the sweet spot for daily usage you are on now to be honest and probably should stick to there until you get new cpu to play with for safety of your components :)
Man... do I have time on my hand today... today is my study day ... (reading book ehehe)
I would guess qx9650 is something about 170w @ 461*9 @ 1.54vcore, this is by testing a qx9650 on a unit I know the performance on.
jinu117
11-16-2007, 07:28 PM
Now, that is a disappointment Tim... that is only on par with Q6600 so far... 170w sounds like the power consumption didn't drop any really either once voltage gets to same level. I assume that stock voltage is much lower? If such is case for heavy volting chip, I can imagine we are back to square 0 of 220-230w CPU heat dump...
yes, I think it is for all of us.
As far as i've seen and heard the reducing of heat by the 45nm technique cpus isn't as much as the gossip have told. the qx9650 is in same watt range as thrusty qx6850 (130w tpd). people say it's a little less hot during overclocking but I'm not quite sure yet since I'm not a overclocker my self and due to the lack of knowledge how to do, I didn't push my qx6850es on SS much before I sold it.
but I fould this on intels webpage;
VID Voltage Range:
qx6850 0.85V – 1.5V
qx9650 0.85V – 1.3625V
VID Voltage range is: The voltage range set by the VID signals as a reference to the VR output voltage to be delivered to the processor Vcc pins.
As far as I understand, the qx 9650 require little less voltage to run in 3ghz (stock speed) then the qx6850.
with other words, I agree with you, we are back on square 1, units still requires tuning for 220-230w.
Regards
Tim
jinu117
11-16-2007, 08:02 PM
Eeek square 1 is what I meant :) Hehe.
The scarey part is... if anyone really remember the jump to 90nm while back, the heat output went up so drastically it wasn't even funny under heavy overclocking. I do believe 45nm should have more lead into top end of overclocking and I can only imagine how much more heat dump that would lead to be. We could possibly sitting and looking at even further heatload (right back to b0 step quad... which some of them I definitely saw 250w at teethering limit of it)
Keep in mind guys.... it seems my wattage numbers are grossly out of line with some trend out there...
http://www.extreme.outervision.com/tools.jsp#cpuoc...
says 315w for what I had fun last time which I only estimated as 200w-230w range... kind of figures...
DarthBeavis
11-16-2007, 08:07 PM
Guys I have a 9650 on a modded vap right now . . .well, it will not boot into windows cause Asus does not have the damned bios out . . .arggg. Go knock on their door and wake em up . . .get me the bios and I will let you know how it runs . . .my phase was tuned for a QX6700 so hopefully it will be fine. . .
jinu117
11-16-2007, 08:53 PM
Which asus mobo is this?
DarthBeavis
11-17-2007, 07:42 AM
asus striker extreme
jinu117
11-17-2007, 10:46 AM
I will search quickly on that.
As for QX9650... I have a customer dropping by with one today so we will get better idea on load :)
word is the 45nm cpus won't work with any nvidia based chipsets because intel didn't get an sli license for x38 (only skulltrail), so it's possible you may have to buy a new board, but first I would make contact with either an intel, nvidia or asus representative to find out if it is necessary
terente0081
11-18-2007, 08:02 PM
word is the 45nm cpus won't work with any nvidia based chipsets because intel didn't get an sli license for x38 (only skulltrail), so it's possible you may have to buy a new board, but first I would make contact with either an intel, nvidia or asus representative to find out if it is necessary
Yes, I also heard that. Not officially, but I've heard that rumor. It sucks because us, the consumers are caught in the middle of this battle between Nvidia and Intel.
If you want QX9650, you have to get an Intel board and no SLI.
If you want SLI, you can't have QX9650.
I want both and I hope a solution will be announced soon. Nvidia and EVGA keep dodging the question, just like they did when they were having problems with Kentsfield support.
xgman
11-19-2007, 08:28 AM
QX6850 at 4400mhz 1.6v
QX9650 at 4800 mhz at 1.55v on a refreshed mod of the LS by Jin.