Wauw ! awsome volts there ! Nice chip Imo
Printable View
Wauw ! awsome volts there ! Nice chip Imo
Thanks mate but its funny how some ppl started to put it on sale on my behalf without me even knowing :rofl:Quote:
Originally Posted by AceBaran
Sorry but you have the wrong idea. I am interested in results which is why i asked you to run IBT or Linx as a screenshot showing 4 or 5mins of P95 is pointless.Quote:
Originally Posted by imo
I take it you never posted this http://www.xtremesystems.org/forums/...18&postcount=7 and you don't have an ebay account with 4 active listings for I7's priced over £300 just because they can run 4.4ghz (not even proven to be stable and have ht off) :rolleyes::rolleyes::rolleyes:
P.s. please learn some english :up:
Either your interested in me more than my chip :yepp:, or your annoyed because you cant afford a chip like this:shrug:,
and yes I do sell chip because I have the capability :cool:
By the way stop dragging conversations made in August I didnt ask anyone to advertise my chip speacialy you.
and talking about English you seem to understand what I am saying so what are you picking at :doh:
Now I can smell something definitely burning :up:
Point proven, your first line makes no sense what so ever..........Quote:
Originally Posted by imo
I may not have lots of benifets like you but i can still afford to buy which ever retail I7 i want (not that this has anything to do with my point and i don't know why you brought it up?).
I just hate people who seem to try and rip people off at any cost (your screenshot and ebay account being a very good example), it really annoys me (so much for a community!)
Maybe i will boot into windows @ 4.2ghz with 1V, run 4minutes of P95 and then throw it on ebay and rip someone off too?
Quote:
Originally Posted by imo
WooW.... very nice Chip....
You obviously miss the point here. Obviously.Quote:
Originally Posted by imo
It's not about who can afford that chip or who can't.
It's about that this 4.2ghz at 1.17V or whatever is NOT proven stable. I can run my xeon at 4,4GHz 1.1V ht on and get into windows, probably run hyper pi 1m for every single thread, but what does it prove? like nothing?
Go run your 4.2G and try at least 10 passes of linx. Let's see how much V you need to add to gain stability ahaha. at the moment it's nothing, but a quick run of prime which is just pointless.
My i7 at 1.17vcore and 4.2ghz running prime bsod pretty quickly... but imo you do realize the screenshot you posted has failed prime... either thread 1 or 2 (not shown) has stopped, since that is how you get an 87.9% load in realtemp, by shutting down one thread of prime.
http://img412.imageshack.us/img412/919/failedoc.jpg
If you look at RaelTemp, it is proven unstable, one thread is locked up ;)Quote:
You obviously miss the point here. Obviously.
It's about that this 4.2ghz at 1.17V or whatever is NOT proven stable.
Oops, rge beat me to it.
Hey guys, I fixed my UPS problem so I began to overclock today. I'm new to the i7 platform (and overclocking in general, sadly) so I am seeking any advice that you can give me.
Here's what I'm currently running at:
http://valid.canardpc.com/cache/screenshot/914217.png
http://valid.canardpc.com/show_oc.php?id=914217
It passed 5 runs of the burn test with 2048MB of memory, max core temps were in the low 70s, is that ok for testing? I know anything over 60 over a long period of time is bad, but I doubt that I'll ever see 100% core load in everyday use.
I'm still working on making sure that it's stable, but does anyone have any suggestions for the voltages? I'm currently using rough numbers from the EVGA site.
http://img205.imageshack.us/img205/6...rclocking1.png
Thanks!
Quote:
Originally Posted by H2O
Utilize 100% of all available memory and make it run 20 times.
You should also try y-cruncher.
The stress testing mode locked up my rig which is LinX stable! O.o
Ok I'll do that. Is low 70s still ok for testing, though?
Yeah, perfect.
Ok, thanks! It's burning now and I'll leave it go while I get my hair cut, I'll report back later.
EDIT: Passed 20 runs of IBT with no problems. On to 4.2!
I have some queries for 920 temps
Ambient temp of 27-28. Using a Megahalem.
What are the normal temps one should get on Air?
@ 2.67Ghz stock and at 4Ghz
There are some who post screenshots of 4Ghz idle temps at 41 for each cores on AIR cooling
while
There are some who post screenshots of 4Ghz idle temps at 41 for each core on Water cooling, so who is actually showing the correct temps or are some people just messing around with the idle calibration under real temp?
Is it possible for a 920 on air cooling to be at 45max on idle at 4Ghz and at mid 30 degrees if at stock 2.67Ghz
Look at this Watercooling dude at over 4gz
http://img205.imageshack.us/img205/6...rclocking1.png
and this guy on AIR. his idle is probably low 30s or mid 40s.
http://img710.imageshack.us/img710/3986/4200mhz117v.jpg
At stock, with eist, c1e enabled, idle will only be a few watts, so you wont see much difference between high end air and water cooling, maybe a degree or 2, and both will be about 7-9C over ambient, during idle.
AT 4 ghz, vcore will vary, and idle wattage will vary, mine idles at about 25-30W if everest reading des sensor is correct, and I idle about 10-11C above ambient. Air will be 1-2C more. AT 4.4 and 1.34vcore, idles at 45W and typically 14C above ambient at idle.
Since there is less variability cpu to cpu at idle than at load in terms of tdp, you can look at peoples minimum realtemp temps and make an educated guess at ambient, but reason you will see such different temps is because of different ambients.
With 27C ambients, you should idle around 35-37C at stock, and probably 41-43 at 4ghz with high end air or water. Load is where water will be much different than air.
Hi, that first shot is mine. I did not mess with the idle temps or anything, I am running at a much higher voltage which is why I have higher temps. My watercooling loop is also not that efficient (GTS240 rad with 4x Slipstream 1200RPM fans, which are not great for use with rads), and since I just setup the loop a few days ago, my TIM has not fully broken in yet.
I'm currently sitting at 4.14GHz (197x21) but I can't seem to break 200 on the QPI. Even with the vcore set to 1.36875, I get an error on the 8th pass of the burn test (1.014922e+008). I pass 10 passes at 197x21 with 1.35 vcore, but setting it to 200 always causes me to fail or crash.
thx for the insight
Finally stable at 4.2GHz. Somehow the vdroop was re-enabled on the motherboard, which caused instability. I'm still tweaking it, but I may drop back down to 4GHz as the cores hit low 90s during the test (I wasn't watching that closely, so I'm not sure if it was a spike or if it was sustained). :shocked:
http://valid.canardpc.com/cache/screenshot/915032.png
http://valid.canardpc.com/show_oc.php?id=915032
I am not against paying any ammount as long as one can afford it, what's important is that these things should bring joy to someones life...enjoy your chip man.Quote:
Originally Posted by imo
Heres my I7. Tweeking isnt over but its still nice result.
http://www.infodupat.com/I7_build/New015.JPG
http://www.infodupat.com/I7_build/150pass_CAS7.jpg
4.2ghz is stable enough for me now. Will start pushing for more (just hope this UD5 doesnt hit a wall) :up:
http://img341.imageshack.us/img341/4148/4220r.jpg
Nice OC Tatts... 4.2ghz @ .992v!!! What batch is that?
Here is my i7 920 Batch 3937A. This 920 sucks but not as bad as the batch 3903A I had previously. This is without Vdroop enabled so the Vcore under load is 1.352v.
Cooling is on water with an EVGA e760 Classified. I don't know what is up with the Gflops, usually linx gives me ~53Gflops.
http://i399.photobucket.com/albums/p...ket/i902-1.jpg
I see the point, which could have happened during capturing the screenshot.Quote:
Originally Posted by rge
So again with a little increase of volt I tested the chip hope it ends all confusions....
http://img109.imageshack.us/img109/3...0119vxreme.jpg
and its not a Giga or Classified mobo for low volt in cpuz!
Yes my point was its a amazing chip to run. I was trying to share my findings but obviously some ppl do take it personally or bring irrelevant discussions or personal bias. Thanks a lot.
Yes this time he sure did...lol:p:Quote:
Oops, rge beat me to it.]
tatts, post a screen UNDEr load. this is without load and Power saving is on. .992v for 4.2ghz is impossible.
@ IMO
We want to we a 20pass LinX with MAX RAM used.
Sorry, this is the load shot just before it finished :)Quote:
Originally Posted by Boulard83
http://img340.imageshack.us/img340/3765/lolsq.jpg
+1 as that still isn't stable :up:Quote:
Originally Posted by Boulard83
Hehe, I know you dont like my comments on some of your posts, but idc.Quote:
Yes this time he sure did...lol
But it seems to be a nice chip you have there, but like many other low voltage week 49 - it seems to be running quite hot with 86c allready at 4.2. Or is your ambient abnormaly high for this test ?
Thx tatts
this little increase made this chip already worse than my xeon, there you go m8, it's not as amazing as you thought.Quote:
Originally Posted by imo
Always glad people share their findings, but you were just teasing, not sharing:p: People are just asking if its stable, or pointing out things that dont make sense.... like I dont understand why your max temps are 85C on core 1, but current load temp is only 72C, 13C swing is bizarre, usually just 5-7C max difference using same vcore/mhz between current load temp and max temp...even when your core 1 had stopped 1 thread on previous only 9C diff.Quote:
Originally Posted by imo
Regardless its a great chip, but was interested in seeing how long prime would run or linx like other people, if you show a teaser, expect people to ask questions or comment on what doesnt make sense.
100% true.Quote:
Originally Posted by rge
Anyone can show a 10min prime or a 10pass LinX.
We want to see a REAL load test for REAL result :)
My I7 920 is very bad.(3911A199)
http://img229.imageshack.us/img229/8927/24916.jpg
@imo
Amazing chip.Very nice voltage:eek:
Has anyone tried benchmarks with and without HT?
I'd like to see how much difference it actually makes.
Nearly nothing if your mainly a gamer.
The only place where HT takes the lead is in the crunching numbers. Like encoding, rendering or similar aps.
here is my new W3570 with big problems i cant use the multis over 24 :mad:
my mainboard DFI X58 T3H6 new Bios,Memory Kingston Hyper X 3x2GB 1800
in bios core1-4 multi ican set
but no non cpu turbo max.24 why?
TDC+TDP is disabled ?? what is this its new in my bios?
i will set 133x35 but i cant this in bios please help me :D
http://www.abload.de/thumb/w357040ghz0h2a.jpg
i have a new batch to test next it's a 3945A936. this is on the new x58-ud7 boad. water cooled...:)
i7 920 D0 Batch 3849A993
@ 4.305Ghz (205x21) HT Off, 1.275v VID, 1.42v VTT
DK X58 T3eH6 0724
KHX2000C9D3T1K3 6GB @2050Mhz 8-9-8-24-1T, 1.68v VDIMM
TRUE120 BE
http://i458.photobucket.com/albums/q...50168vdimm.jpg
Thats a big jump from your original 1.17V and you still havn't done 20+ passes :rofl::rofl::ROTF::ROTF:Quote:
Originally Posted by imo
.
+1 :)Quote:
Originally Posted by tatts
Will disabling HT allow me to drop the vcore a little bit or will it only lower the temps?
Will probably help with VCore as well, yep.
But do you not want to keep HT?
I am testing the same batch, 3945a936. So far it is LinX stable 4ghz 20 passes, 1.248vcore load, 1.220vtt. Testing it now at 4300mhz 1.360vcore load. The VID is 1.28125v, kind of high but temps are great.Quote:
Originally Posted by top game 22
http://img23.imageshack.us/img23/1892/linxcas6.png http://img23.imageshack.us/img23/lin...png/1/w967.png
oh guess it doesnt make any sense but I dont see any teeth in those smiles lol :D
We just want to see something that really show its stable.
10-20 LinX pass isnt "stable" at my eye but its a beginning !
I don't have much use for it - I do no heavy photo editing or video encoding nor do I participate in things like F@H. My i7 rig is mostly for gaming and (ASAIK) none of them support more than 4 cores.Quote:
Originally Posted by zalbard
Batch # 3939B145
Here're my settings, finally hit 4ghz. So far 5+hrs stable in prime with decent load temps.
http://i195.photobucket.com/albums/z.../Capture-1.jpg
Hyperthreading is enabled. Mem set to 8-8-8-24.
System:
Gigabyte ex58-ud5 board
Core i7 920
Corsair Dominator-GT 6GB DDR3 2000 @ 1600mhz
850W PSU
gtx 295
watercooled, 2 fan rad.
I was wondering if you guys think I should go for 4.2? I am was going to enable turbo mode, which will give me 4.2ghz, but I am kind of enjoying this somewhat low volt overclock.
And if I am using turbo what will be the voltage bump i should go for.
I do nothing, but game and surf web on this machine. So I think even 4ghz is overkill, but to me there's no point building almost 2G machine and not push it.
My latest high BCKL, low-volt, low-VVT, 24/7 setup. ;)
VVT=1.28v :)
W3520 @4.2GHZ(19x221) 1.20v :cool:
http://img.techpowerup.org/100101/Ca...3-20100101.jpg
Comming from R2E.. now X58 classified (much more stable, at least for me it is)
http://img706.imageshack.us/img706/1791/45stable.jpg
Finally you are able to break the 210 bclk on the R2E....with a classified :up:Quote:
Originally Posted by AceBaran
If i was you, i would use linx or IBT, prime takes too long and that worker3 looks like it is struggling to me (mine did the same for 6hours then failed ibt within 5mins lol).Quote:
Originally Posted by Davit
4.2ghz shouldnt take much of a push though :up:
Such a releave :pQuote:
Originally Posted by Salamndar
I have not received mine yet....Holiday season is screwing with my shipments.Quote:
Originally Posted by AceBaran
I can't wait to try my 950 Batch# 3931A332, so much 3s in that batch !
I have a question for the guys that run Prime Blend. Are you running it custom with more ram used or just default? Default only uses 1600mb of ram to test, it was suggested that I run custom using 5000mb's.
Thanks
Thought I would give it a try, I set blend to custom 5000mb. Will let this run over night.
http://img51.imageshack.us/img51/938...lend5000mb.png http://img51.imageshack.us/img51/cus...ng/1/w1493.png
3938b020 ........... 4.2ghz 1.2v......4.0ghz at stock 1 ghz.........screenie's soon, testing stability.
I actually did.Quote:
Originally Posted by tatts
After readin the forum, that is what I used.
Haven't pushed her to 4.2 yet, but with my temps being low I might.
since i passed the linx I am gonna try 4.2 at the same voltage... see if I can break anything lol.
Water installation done. After 10 mins playing with bios. Classified kick ass :clap: Will try to either push more bclk or reduce vcore a little.
Eleet shows wrong voltage, dont count on them. Bios vcore 1.3125, vtt +50mv, dram 1.63v, qpi pll 1.20v
http://i49.tinypic.com/b8tlle.png
Nice work guys.
Just a reminder to please put your overclocks into the Xtremesystems database I made (make sure you show stability test in SS) - found Here
Add your OC details to the form, then click submit and the database will automatically update with your results. There is a link to the database in the thread I linked you to and
you can download it yourself in various formats as you wish.
Common guys, we don't want other website forums to be more organized than us do we? :)
Can we update our entries as we tweak the overclock?Quote:
Originally Posted by CryptiK
Absolutely, they all count. Just make sure its stable with a stress test then add the results to the spreadsheet.Quote:
Originally Posted by H2O
Oh, OC Nub - most people just run a default settings blend, but those more serious about it run custom blend with all available ram used as it is more stressful.
wow 4240 is the highest thats been posted so far. I was looking through a substantial list on another forum around the first of Dec and there were several people up in the 4.5-4.8 range on water. Most of the higher clocks though I believe were on low B batch chips.
I'll start overclocking my 950 Batch 3931A332 as soon as I get my New ZM1200M power supply.
The database in first post in this thread has not been updated in nearly a year, and is with old C0 chips.Quote:
Originally Posted by scgt1
Here is more recent thread with D0 i7s in 4.5-4.8 range
http://www.xtremesystems.org/forums/...d.php?t=239175
It is interesting to see difference between C0 and D0 overclocking...
Threads where the original OP has to keep updating and organizing almost always fail. That's why I made the self updating database.........
Here's a low voltage low temp clock.
3.6 GHz @ 0.992v | 1.15 Vtt | DDR3-1715 | 8-8-8-24
http://i720.photobucket.com/albums/w...DDR3-1715M.jpg
http://i720.photobucket.com/albums/w...-DDR3-17-1.jpg
very nice temps CryptiK :up:
Cryptik, you should lower your multiplier and raise the bus speed to bring QPI link speed up to the core speed for optimum performance.
Thanks Radaja :)
Mngdew - I was just playing around. I have a very hot CPU and I was seeing what it would take to keep it around 60*C load. Usually run at 4 - 4.2GHz. Water coming soon so I'm doing some 'before' testing.
Thank you. I have become lazy in my stability testing but still like a long blend run for a final test.Quote:
Originally Posted by CryptiK
Mine also runs hot. Even with the water cooling, the first core reaches up to 80C during LinX runs under 24~26C ambient.Quote:
Originally Posted by CryptiK
i7 975 EE
batch: 3910A347
4.2GHz w/HT ON @ 1.265V
http://www.pohrani.com/t/v/bV/3ReUsBx3/asd.jpg
Max frequency under LN2:
http://valid.canardpc.com/cache/banner/928917.png
3D benchable frequency under LN2 w/HT ON is 5173MHz with 1.55V vcore and 1.45V VTT.
This was all done on Asus Rampage II Extreme. I think that on EVGA Classified this processor can go more.
920 D0 3926A446
4.2Ghz (HT Off)
Thermalright 120 Ultra Extreme + 2x Aerocool Silverlightning @ 1300rpms
http://i770.photobucket.com/albums/x...2HToff1329.png
Still testing...
CPU:
920 D0 SLBEJ 3903A468 (purchased June 30th 2009)
4.2ghz (200x21) HT OFF
1.2875 in bios (1.325v with LLC off), 1.264v under load
1.34v QPI/tt (<1.32 and >1.36v cause 0X00000124 BSOD's)
1.24 QPI PLL
stock 133x20 does 1.00v in bios on auto and is stable, but needs 1.2v QPI/Vtt. i take that as good vCore binning.
Cooling:
noctua nh-u12p
http://i52.photobucket.com/albums/g2...4P/42Prime.png
core 4 always reads 4-6C cooler than cores 1+2.
as stable as: 48 hrs Prime 95 Small FFT + 48 hs custom Blend + 24 hrs Linpack (left running and went out+forgot about it :zombie:)+ 24 hours furmark+Prime blend all without rebooting.
been running for 6 months with no un-expected program errors/crashes.
http://i52.photobucket.com/albums/g2..._0226_800x.jpg
Have you tried to get 4.2GHz with HT enabled? What vcore do you need then?
Same question here, I can set 3800 (200x19) at 1,20V Full Stable but 200x20 or 200x21 not stable at 1.3V+ with HT ON (+0,1V jump and it isn't stable). With HT OFF no problem at 1,2825V. What vcore do you need to 4.2Ghz with HT enabled?Quote:
Originally Posted by madmax999
You have to do some testing to find out. Not all cpus are the same. I can do 21x184 HT on @ 1.2v but i need 1.33v to pass linx at 21x201 HT on.Quote:
Originally Posted by r0ckS0LiD
I haven't done any extended OC'ing with HT on past 4.0ghz (190x21) because I use this system for gaming and the heat output was getting outrageous. IIRC at 190x21 i needed 1.28vcore with HT ON, where it only needs 1.22 w/ HT off. The QPI/Vtt also needed 1.42v w/ HT on to pass several days of stress testing, although it might have needed voltage elsewhere. I don't have the need for HT so I probably won't ever go back to testing with it on.
920 D0 3926A446
4Ghz (HT Off)
Thermalright 120 Ultra Extreme + 2x Aerocool Silverlightning @ 1300rpms
Vcore: 1.258v
VTT: 1.28v
SP2004: 9h (priority 10)
http://i770.photobucket.com/albums/x...real/th_9h.jpg
i7 920 Batch # 3845B259 or 3844A953
i7 920 Batch #: 3927A516
http://image.ohozaa.com/is/4.5memall.png
http://image.ohozaa.com/in/4.5memall1.png
i7 920 batch 3849B041 on water using a GTZ block with a PA 120.3 rad and MCP655 pump. This is my 24/7 speed.:D
which one should i buy ?Quote:
Originally Posted by Dom7184
Grab the 3845B259 as it should do pretty good. That is the start of the good steppings, 3845-3849.:cool:
will its sold :( wut about the 3945A854 ? any good ?Quote:
Originally Posted by flesheatinvirus
I honestly do not know about that one though most can do 4G pretty easy with the right combination of hardware. If the price is good pull the trigger I think you will be happy.
will should find out thursday it gets here but ill be at work :( so friday ill get to hook it up, idk if i should try water or phase 1st :p: thinking water just to make sure everythings working lolQuote:
Originally Posted by flesheatinvirus
this is wut else i got Enzotech SCW-REV.A CPU Waterblock, ASUS P6X58D Premium, G.SKILL Trident 6GB (3 x 2GB) DDR3 2000
Cool hardware setup that should work out real nice, Enzotech make some nice stuff. I too would start with water just to iron out the bugs then phase it after you figure out a 24/7 and bench speed setting. Gskill Trident is going to love that setup. I am running there F3-17066 6GB set on this P6T and they can do 2245 with some juice or 1720mhz at 6-7-7-22 at 1.64v. Some really nice sticks for sure.
I7 920 Batch 3945A781 4.2GHz 1.28vcore with air cooling
3938b020, megahalems pushpull, 4.2ghz 1.192v HT OFF, HT ON 1.208V , 24/7 stable
http://i749.photobucket.com/albums/x...lentsnip/1.png
http://i749.photobucket.com/albums/x...lidation-2.png
The link is broken it seems.Quote:
Originally Posted by melnd
Haven't messed around with my i7 lately what seems to be the max QPI people are getting?
Finally got my hands on one of the "coveted" 920's :D, Batch 3845B259. While it is a bit better than my W3520 was, it's not earth-shatteringly so.:shrug:
Notes in pics give voltages.
4200mhz
http://i213.photobucket.com/albums/c...X20xstable.png
4300mhz
http://i213.photobucket.com/albums/c...X20xstable.png
4400mhz (sry, forgot to show memory timings in CPU-Z.)
http://i213.photobucket.com/albums/c...blelessqpi.png
Still working for 4500 with HT on. Voltages are starting to make me a little nervous though.
I am happy with my 3849B041, just getting started.:cool:
I got hold of a retail i7 930 today wich will be in my hands tomorrow :D
I can already give the batch : 3001A117
Results soon :)
Very nice Smokey. Looking forward to your results.
How the hell did you do that?Quote:
Originally Posted by SmokeyTheBandit