Yeah, it's safe, but it's pretty hard to get stable with such temps from my experience.
Printable View
Yeah, it's safe, but it's pretty hard to get stable with such temps from my experience.
I ll give it a shot and report back. If i will not manage to get is stable at 4.5 ht on then you will make it a 4.4 club :p:
The only mod i can make is to open the window at night to get some 10C cold air from outside and close the heating, LOL
w3520 3945a Water cooled with the help of nature. Ambient temps between .5 - 1c. Everything is overvolted more than it needs to be but I failed LinX at 1.5 Vcore after 16 passes and it was getting late so I just cranked up all my voltages ftw.
My 2 previous i7-920 39xxx batches couldn't come close to this so i'm pleased with this Xeon.
http://i399.photobucket.com/albums/p...520-45pass.jpg
Updated.
@ trn: good job, and very nice temps! :up:
I would like to join the club too :D
http://img25.imageshack.us/img25/6195/45001hlinx.png
@alfaboy, very nice! Do you know what batch your 920 is?
could linx it but summer now so i will show blend...running 2,5hs+ on a winter day
http://img64.imageshack.us/img64/5633/45001200.png
Cheers
Sergio
Do i5-6xx count for this? :p:
If so, that list is going to grow here pretty quick.
awesome party you have here, I wanna join :D :party:
http://img297.imageshack.us/img297/2408/4500cut.png
aka_GK
Very nice :up:
What are the odds of the list being updated with batch #'s?
A lot of people don't want to remove their coolers just to tell us their batch numbers so this will not make it into requirements.
A few people listed them, though.
But there are so many batch number threads already so it shouldn't be a big deal anyway.
However, if people wish to list their batch numbers, I will surely add them to OP.
Well, i got this monster on water. :rofl:
test on water rig:
ddc3.25/xspc res top->He120.3(3x GT-ap15)->HK 3.0 cu.
ambient is 20C.
Very disappoint the Evga board not supports this CPU now. still waiting the new bios from EVGA. :(
http://i645.photobucket.com/albums/u...g?t=1264333159
more update,
4.83G pass linx 50 run on water :rofl:
http://i645.photobucket.com/albums/u...g?t=1264515192
Nice results! Looks like cherry picked retail Gulftowns should be able to get to 5GHz on water! :D
Yes, The 32nm chip very good on the temperature.
This chip can do 4.6G pass linx 50 run on AIR.
Just under Megahalem with GT1400RPM :D
Where the hell did you get that chip at? I can't find 32nm Xenon chips at the egg. I did a search for intel 1366 and came up with nothing but the older 45 nm process chips.
Yo Zalbard, i have 12gb of memory. For how long should I run the test to join the club? )
Would be around 2 hours for 20 passes unless your i7 runs @ 6GHz. :rofl:
If you can't manage that long then your system is not stable for the used settings I guess... Try chilled air or pulling out 3 sticks of RAM (assuming you're using 6), such kind of 'cheating' is acceptable. :p:
Chilled air is the best option I guess, considering the temps outdoor ))
I would expect a 2nd retail stepping of the 32nm process to be monsters. Kone's OC's are very nice but this is an A0 stepping ES. I would speculate that Intel still has more to learn and improve the 32nm (in hexa and quad versions.) Its my own opinion that future 32nm retail steppings will be excelent (based on my own wild-ass-guess!) :D
Oh... and awesome temps for that voltage Kone! :up:
Very good results i could not pass 4.6ghz, your results are great....
Also i am sorry to tell you that the temp reported by ur cpu maybe wrong. These dont show the correct temp and i have been told it will get fixed in B0/B1...
Also i will soon get a B0/B1 and give this test another try... :)
hi
screen shot after a hour of occt
Hey. :)
This is a LinX stable club, I'm afraid OCCT won't cut it. However, they are very smililar, so I am pretty sure you can pass LinX with the same settings. :up:
You can find a download link for LinX in the first post of this thread.
And please do not downscale your images. Makes them hard to look at, keep them at 100% but crop off the unnecessary bits to reduce the size. :yepp:
Good afternoon,
i7-965 at cpuV of 1.6V, cooled by 2xCoolit Boreas in series with GTZ block - c0 stepping needs lots of volts!
http://img192.imageshack.us/img192/1926/45i7965.jpg
This the same mobo, cpu and mem that has me at #2 in the HOF for pcmark vantage -
http://www.futuremark.com/community/...pcmarkvantage/
sorry for the offtopic: what the difference between Intel® Core™ i7-960 and Intel® Core™ i7-965 Processor Extreme ?
Good evening, Disregard my previous this one is better - slightly higher clock, with voltages (easytune6) and more memory/longer linx run. Still chilled by 2x Boreas in series with GTZ block. -
http://img693.imageshack.us/img693/8827/45betterz.jpg
4.5GHz ES, 12GB (3x4GB GSkill Ripjaws), P6T Deluxe V2
Air Cooled by Xtreme Gear 1216 B with Y. S. Tech 8.52W fan (Standard XG 1216B fan 95C)
Vcore 1.625 V applied, 1.568 V CPU-Z
http://farm5.static.flickr.com/4015/...0eea5b30_o.gif
I don't recommend these settings for 24/7 use!
Clarkdale next, surprised I haven't seen more of them, 4.5GHz is easy there, they'll do 4GHz on a stock cooler!
215x21 and 25x180 are good for 4.5 but I couldn't either it LinX stable, probably due to using 4GB RAM modules or not being enthusiastic enough with the voltage.
1.65V was too much and had temps pushing into the 90s, 1.6V wasn't quite enough.
The only place I've really seen odd Multis making a difference is 191x21 vs 200x20 for a 4GHz OC on i7.
Holy crap, damn right I wouldn't :eek:
I had a Westmere-EP die on me at ~1,65V.. on SS and no I wasn't running prime or LinX :rolleyes:
Here's to hoping that my X5660 B0 can handle 4,5G with 1,4-1,45V max (it should) - just need DFI to release a bios that works for B steps...
Or evga to make one for Westmeres, for that matter :rolleyes:
Quick question about the reqs, I know that LinX and at least 2GB are required, but what about the ratio on which the RAM is running? I ask because I'm currently forced to use the newer DDR3 RAM [1.65v] and it does not come close to being able to run at 1900Mhz eventhough it is DDR3-2000 RAM. So, my RAM is running at 1:3.2 Ratio with my 450FSB which equals to 1441Mhz. Let me know if that is not an issue so I don't waste my time looking for/finding the LinX program and running the 20+ passes.
Don't care about your RAM speed as long as there's at least 2GB installed!
Ok, SteveRo's PC has 4GB of RAM and it shows it took 129.xx per round; I'm guessing these are seconds. So it should be less than an hour. Thanks.
Gulftown - 16 March - I'm going to get one if I can.
Nice thread Zalbard :up:
Had to lap one of the i7 to get my name registered in your club lol :D
Before
http://img52.imageshack.us/img52/3804/ccccs.jpg
After
http://img33.imageshack.us/img33/4747/lapy.jpg
Thanks mate, well what can I say I was inspired by the club and fellow extreme clockers here, yes very nice chip to clock, the main problem was heat , lapping helped about 4-5c drop on load beside the ambient was low aswell about 16c or something lol I love winter ;)
imo
Great job, man :up:
i7 920 Batch #3938b024 1.3750vcore ht on @ 4515Mhz Swiftech apogee xt water. Still working on 4.6 :)
http://www.straferight.com/photopost...20_4515Mhz.jpg
I'm pretty sure I can do much better than 4.5Ghz, but right now I'm remodeling my house and the only "clean" room is my bedroom and it is hot as hell in there [wife got the heating system in near full blast]; also, I'm playing around using 1.65v DDR3 RAM, that's the reason behind the low RAM speed of 1442Mhz. So without further delay, here is my PC, 4.5Ghz w/ 4GBs of RAM [used 2GB in LinX since it only had just over 3.3GB available] Cooled by WC: DTek CPU block, stock BlackOps NB block, EK HD4870 block, ThermoChill PA120.3 w/ 3 120mm Scythe fans, external 500cc aluminum reservoir, 2x Swiftech MCP355 [I think] coupled by an EK dual top. System BIOS settings: CPU 1.34v, NB 1.40v, VTT 1.25v, RAM 1.64v [my board jacks it up a bit, so I got to fine tune this later]
PS: Let me know if you need me to run LinX with any different settings; but the system is rock solid, as you can see it passed 28hrs 40mins of Orthos before I paused it to do the LinX test.
http://img208.imageshack.us/img208/5...nx45stable.jpg
Very nice, I'm jealous of such low vCore chips! :D
I've added your data since it doesn't contradict with the rules, but I can just note that the more RAM LinX uses the better, it increases the stress level, so the next time testing for stability rather use the "All" button to use some more of the available RAM, you could've pushed to 2800 or so I think (out of 3300).
Here is the LinX test done with 3GB:
http://img246.imageshack.us/img246/6...x45stable2.jpg
Here's my 24/7 3849A895.
--Matt
let me join
i7 920 Batch #: 3927A516
http://img.zuzaa.com/image.php?id=29D4_4B7019AE&jpg
http://img.zuzaa.com/image.php?id=2EC3_4B7019DF&jpg
I just ordered a UD7 and a 3520 - will be putting it under water with apogee XT and 2x boreas in series.
Looking for 4.6 or better by this time next week- we'll see.
I figured since my original post was done before the required 20+ passes was set I should show another run. Its the same chip just done on my DDR3 setup this time. batch is L827A945
http://i131.photobucket.com/albums/p...600c6linx2.jpg
Thanks buddy:)
Thanks! The lapped i7 batch was 3849A827 out of 2 other same batch this one is cooler and lapping made it coolest :cool: yes the Xeon also does 4.5GHz+ @1.37v One thing I have noticed that same batch CPUs overclocking result is not same even on same setup, Therefore its true "every chip is unique" :yepp:
http://www.f-thirteen.net/adam/LinX20_4600.jpg
Air-cooled with a lapped and washer-modded TRUE.
this thread makes me drool with jealousy... lol
Does my Xeon with just 2C/2T count? :D I can make it 220x21 stable on AIR, but just 2C/2T. Not anymore :)
Nope, I'm not using it 24/7hahahahah, not even 10% of time on my computer :) Just for spi, but in past It passed 5 runs of linx with max ram used - I think it can pass 20 runs w/o problem :)
Picked up an i5-660 yesterday and put it in my P55A UD6. Used 2 sticks of a 3x1gb triple channel kit of Kingston HyperX from my x58.
This was the easiest 4500mhz+ I have ever done, hope this chip doesn't end up on Jor's i5 Graveyard thread.:yepp:
Thank you PangingIII for your help with qpi settings. :up:
http://img697.imageshack.us/img697/7...zi5660cas7.png
adadk
impressive temps for air :up: Great chip also :)
plz add me, thanks. 4.510ghz 1.34vcore....24/7
http://i749.photobucket.com/albums/x...tsnip/1415.png
There is some very impressive overclocks here!
@meldn
in your screen i see that you've got hyperthreading disabled... What is the max 25 passes linx run with hyperthreading enabled of your chip???
Good morning to you all,
This was not a particularly good W3520 based on what I have seen from others.
I raised cpuv all the way to ~1.65v and could not go any higher than what you see below – please add this to the list, thanks.
Also FYI, this is the same mobo, mem and proc that I used to establish the new PCMark Vantage WR –
http://service.futuremark.com/search...ingsystem=-100
http://img694.imageshack.us/img694/2905/45w3520.png
^^Your chip may not be golden but your cooling sure did a great job. Nice run.:up::up:
For what I spent I would have been better to go with SS.
2x coolit borias in series with a apogee XT block.
A little better than straight water but only marginally so, see -
http://www.xtremesystems.org/forums/...d.php?t=217758
More specifically -
http://www.xtremesystems.org/forums/...4&postcount=63
add me plz :))))))))))
4.510ghz 1.34vcore!!!!! 1.35 qpi, 1.86 cpu pll, 1.14 IOH, 1.1 ICH
http://i749.photobucket.com/albums/x...p/4520pass.png
http://i749.photobucket.com/albums/x...nip/121415.png
http://i749.photobucket.com/albums/x...snip/turbo.png
melnd,awesome cpu,plus a great OC with low vcore.do you have HT turned of because of temps?
i bet if temps permit it,you could do that same run with just one notch up on vcore w/HT.
3938B020 920 4.510ghz 1.34v(bios) - CPU
it's the one in his sig i think
Thanks, Guess I should have seen that, oops.
Very nice cpu melnd... :up:
Just wc this bad boy... :D
There is something I have always wondered about...
45nm i7 s1366.
Once you're stable at low multi and high bclk (let's say 15x220), could increasing the multiplier (let's say to 21x220) require a higher value of CPU PLL and IOH PLL (1.8V for 21x instead of 1.5V for 15x) in case they are undervolted? What are your observations? I cannot seem to find a pattern, my results are inconclusive.
New radiators, testing heat load. Alphacomputer 360XT+240XT with 10xD12SM wasn't enough. Now PA120.3 and PA120.2 with 10xZalman F3, 7c lower idle 12c lower load. Could be more in it, chip hates bclk tho and board unstable 192-199bclk. Takes vtt@1.53 to do this bclk with turbo off. Stable @ 4.565 with HT on, I find it faster gaming with HT off though at 4.675ghz (and cooler) so thats my submission :up:. May be an update in the future if I can find the time to push it some more.
http://i169.photobucket.com/albums/u...4675stable.jpg
@ PiLsY: Very impressive! You seem to be lucky with your chip, so many people are not fond of their 950s and 960s (and 960s seem to be even worse).
Welcome to the club! :D
Good morning Mr. PiLsY,
Wow!
Well done, excellent OC - very impressive indeed!
More to come, just need to find the time to do it! That was a run at the same clocks and volts as my 24/7 setup on my old rads to get an idea of the temp difference. I hope I can drop the core volts now so I can drop vtt too. This cpu needs vtt to be close to 0.01v over vcore to stay stable at high clocks. As you can imagine this means temperatures ramp up fast when you add voltage. I've noticed this tendency on all recent D0 cpu's ive had through my hands (last 5 920 chips all made in '09), the earlier '08 D0's were happier with vtt between 0.3v and 0.1v lower than vcore in my experience.
Left it running 24x200 prime blend while Im at work same voltages. I expect it to fail, I just want to see how fast. On my old cooling it could do about 40 minutes. Prime only generates max temps of 67c this time - again approximately 12c lower than before.
CPU details:
3910A470 vid:1.1875v