With HT Turned off Games Actually perform better.. and pretty much all benchmarks are higher
I sure do like the - 10C cooler
I am thinking about doing dual loops with mine and thursday
feser360 for cpu
pa120.2 for GPU
Printable View
With HT Turned off Games Actually perform better.. and pretty much all benchmarks are higher
I sure do like the - 10C cooler
I am thinking about doing dual loops with mine and thursday
feser360 for cpu
pa120.2 for GPU
I have my 965 @ 4.275, 1.45v, HT disabled
I get 64c under full load at these settings. If you'd like, I could do a prime run at 4.0 with HT enabled and report back my temps...
And you're totally right about performance w/out HT. I get at least 10fps higher in all my games with it disabled up to 45fps in a few :)
sono.. yeah turn it back on and tell me your temps
So i guess running core 920 at 3.8> GTX M184 92mm x2 HWLabs Rad> (3) GTX280's EK > GTX 480 HWLabs Rad> Swift MCP655 Pump just ain't gonna cut it? Son of a .......
480 running SCYTHE S-FLEX™ "F" 120mm 1600 rpm Quiet Fan
:DHT OFF and the cpu is stable @ 4GHZ -1.375 volts and temps lower by 10c.:D
Prime 95 over 1 hour, going to leave it over nite and see if its good for 8 hours.
http://i28.photobucket.com/albums/c2...enom/htoff.jpg
These i7's really are proving to be hot chips. Hoping newer revs will make a diff. Like QX6700 B1 -> B3. :)
core i7 is the hottest CPU i've ever tested
i reckon they are on par with those power guzzling Pentium from a few years ago if not worse
But it seems to take it well. To awnser the OP I just got my i7 920 hooked up under water last night and immediately overclocked it to 4.0 ghz. Right now i'm idling at 45C and full load levels out at about 75C. I have a Swiftech Apex Ultima Water Kit in a Cosmos S. My Rad is mounted to the to (inside the case). As you know this kit comes with the Apogee GTZ Block with 1366 Conversion Bracket which seems to work very well.Quote:
Originally Posted by dinos22
My voltage is set to 1.425 with HT Enabled. I'd say I have a little more head Room to get 4.2 or 4.3 out of it. But I have'nt tried it yet. I'm satisfied with this overclock right now and it's my 24/7 as of right now.
Guys I re-read my post and it sounded like I was moaning after I re-read it. I am not crying about it or anything. I just was kinda shocked, and want to get the best cooling, and do not wanna spend money until I figure these out a little better. I'm not gonna give up or anything drastic like that. :)
I would say it has to be better than water cooling to run this thing. I'm using True copper lapped and temperatures are like 100c load 8 cores prime 95 1.40V full load, prime 4 cores no problems at all.Quote:
Originally Posted by T_Flight
Phase would do wonders to contain the heat of this CPU.
Hello Guys,
I cannot believe that decent watercooling, i.e XFC 240 rad, something like Fuzion v2 would not be able to cool these i920's down. Me thinks it's an issue with getting the heat to escape from those hotspots inside the cpu, which seems to be getting worse with higher voltages - maybe a similar situation to Prescotts of yonder?
I would luv to see you guys use a simple XFC 240, and compare water inlet and outlet temps with a XFC360 or XFC480 for example. I would like to bet that the cpu won't even come close to saturating the capcity of even the XFC240 rad.
Would be nice if you guys can test this out. I am running 2x Xeon 5430 Harperdales, and a 4870X2 with decent results on a single XFC240 - so the results you guys are posting doesn't make sense, and I think that even throwing 2x XFC480's in the loop won't change much, as it seems the heat gets trapped in the cpu somehow.
Keep us posted.
MrBean
you are probably not far from truth
i'm sure that block design could use some tweaking for these new hotties :D
@ Dinos22: Yep, better tweaked blocks might help a little - I think Intel will sort out leakage currents pretty soon in next few steppings - we can only hope :)
Ah poo, now I don't think I want to go i7 for my project.
Maybe wait till Phenom II and see how they go.
Wait wait: Maybe it's something as simple as a waterblock revision, that help to transfer the heat.
So far all the blocks are for other sockets bar, 1366.
Then again, it could be intel that need to sort out leaking current and get some revisions out soon.
3.9gz
1.375 v
36 idle
62 -65 load
HT turned off turbo enabled
d Tek fuzion v1 with quad nozzal and spacer
thermocil p120 & p240
hd4870 with EK water block
ddc2 with xspc top
3/8 tuding
Mcubed T balancer controlling fans & pumup
only problem is no back plate or fitting for my fuzion block
so have temporary solution
wondering wether its worth changing to swiftech gtz block
temps seem ok and totaly stable
These things may seem hot but they run well at those temps.. remember when the 4870's came out? everyone was freaking out about how hot they ran. I think the temp issue has to do with the HT technology putting more stress on each core. But like i said, it seems like they've put this chip on higher grade silicon with better soilder and circuitry bonds because the heat does'nt seem to bother it at all. 100C is the new standard not to go over with these babies. With the Core 2's it was 70C. You guys just got to get used to the chip i guess. Just like everyone got used to there 4870's running at 80C
Quote:
Originally Posted by beast200
It seems like you have a good chip but you dont have HT enabled and i have to believe that takes some of the voltage requirements and thermal output down.
hmm 2 pages, not a single ambient temp post...
all your ambient temps most likely are in the 35C and above...
someone with 20 to 25C ambient please post some i7 temps...
Wow, I'm amazed (more like shocked) with the high temps :shocked:
And septim is absolutely right. :up:
Please post delta from ambient temp (room temp) of your i7 or at least the ambient temp.
I think measuring the water temp can tell us if the blocks are able to extract the huge amount of heat or its still stuck somewhere inside the chip which is the cause for the high temps.
Can anyone measure the water temp during a load/idle of an i7 @4Ghz (or close to that).
BTW, thanks for those who share their info.:)
Ya... I have a SS. Unmodded Vapochill. When I 1st got me E8500, which is a poor sample really, I used to say 'baaah nehalem will be the time for the SS' as the E8500 needed silly volts for anything close to 4.4Ghz.Quote:
Originally Posted by Metroid
Of course my SS is rendered useless again with the power heat output from i7. Almost thinking of selling it TBH.
T-Flight... just think... if you don't have water, you may not come close to the clocks reported here.
when my new stuff get here tommorow I am going to see if I can tame my Core i7
I am going to run Dual Feser 360 and a PA 120.2 and MCR 120
my room temp on all my test have been 69F my computer room is cold
also I got a probe to monitor water temp coming tommorow also
water temp don't differ too much. its air in air out that's what important.
comparing temps, you should have your ambient temps posted always... it will be the basis for temp comparison... as in how high from ambient is it at idle or load...
this would be assuming you have proper airflow inside your case, any case tower, cube, htpc.
with this in mind, there won't be that too much factors limiting a comparison...
What temp do the i7 920 throttle at?
so I did a prime run @ 4.0, 1.36v, HT enabled
this is on water with a GTZ, mcp355 w/ petras top and thermochill 120.3 w/ 6 fans in push pull config on the front of my MM case sucking cold room air
ambiet is around 75f
http://i37.photobucket.com/albums/e8...ander/40ht.jpg
for reference here is a prime run at 4.25, 1.46v, HT disabled
http://i37.photobucket.com/albums/e8.../4250prime.jpg
I quite agree, need more ambients. :yepp:Quote:
Originally Posted by septim
Also, aren't you guys forgetting that there is something else that's new about the Ci7 that also generates more heat? ;) ;)