Oops! Guess I spoke too soon. :DQuote:
Originally Posted by barzi
Printable View
Oops! Guess I spoke too soon. :DQuote:
Originally Posted by barzi
What operating system?Quote:
Originally Posted by barzi
And what programs and software have you installed since the first successful test?
A bunch of software. Begin was with a clean XP install. Now I have installed games etc.Quote:
Originally Posted by pH(x)
OS = Windows XP SP3
But what should I do: 1.288v for completely stable(orthos too), or 1.264v for game/usual things stable? Only orthos failes, games etc. don't fail.
And is 1.288v a safe voltage? So definatly no degrading cpu?
Given a couple degrees difference, my e8400 exhibits wide variability. At less than stable settings, it also exhibits variability at same temp. If you let it cool off some, ie turn off or idle for awhile and/or ambients drop or test on another day, you might find you can run it longer again.Quote:
Originally Posted by barzi
But I think you are just showing typical variability of short runs of orthos with these E8xxx's +/- temps playing a role.
If you get it orthos stable at least 6-8 hours, and you can go to custom and set 8 to 64 small ffts in place (test like newer version prime), then on subsequent runs, while it may crap out at 3 to 10 hours or not, its highly unlikely to stop at 3 mins. Unless you have a defective CPU.
You may find that at temps less than what orthos produces, you are perfectly stable with less vcore than is needed to run orthos stable. Though in your case, stopping at 3 minutes, I would raise vcore, and at least get a few hours stable, consistently.
Hi
After much time has come my E8400, my first result stabilising 4Ghz with lower voltage possible
http://img99.imageshack.us/img99/915...hz1248vts6.jpg
What cooling are you using?
Cooler BoxQuote:
Originally Posted by ThatGuy16
A musing: are the instances of degradation/dead chips resultant from over voltage and electron migration, or are they due to temperature? Most comments in this thread, especially initially, were that degradation and migration would take significantly more time than the matter of days and weeks from whence issues have arisen for people here. Since the chips seem to be so temperature sensitive, is it possible that the higher temps reported might actually be real, and hence the processors are essentially cooking themselves at anything over stock? Again, just a thought, but I'm in the process of tweeking my own E8400, and I want to think about what I do before I screw up my own processor. I also should think about another cooler since I'm using the Intel stock cooler :)...
>1.6v PLL, >1.2v VTT (fsb term), >1.365vcore, >62c
Personally, I'd avoid doing any of the above with 45nm. xVenix, you have a point, but it's still kinda early to say. I will say that I have heard from reputable source that high pll is the culprit on some of the chips he's seen fried. It's a problem if people only consider vcore and temps.
Funny, I keep getting errors on Core 0
I ran Prime95 last night at 8x500 on 1.29v and core1 was fine, but core0 crapped out after about 30 minutes.
This morning I left it running with 1.31v (bios) and this afternoon when I got home, same thing... core0 crapped out after 30 minutes, core1 is fine.
zlojack, you might want to try adjusting the GTLREF for that core up and down a few percentage points to see if it helps with stability on the first core.
I could definitely see high PLL voltages being a problem, it has been practice with the 65nm chips to increase the PLL voltage before the core to attain a stable overclock. Perhaps you need alot more voltage on PLL over 4ghz, but I have it currently set to 1.17 and I'm seeing 1.18 in hardware monitors.
I can't see how people can do assumptions that they have STABILIZED their system by running Prime95 for 30 mins or so..... omg....
When I was in test phase, it happend for many-many times over the years that orthos/prime gave me an error after 5+ hours of testing....
So pls, dont do this, lets just say, u MIGHT have a chance to make things stable...
If it is stable enough for them to use daily and no crashes who cares??????Quote:
Originally Posted by laragirl83
I think you should take your car out and run it as hard as you can for 12 hours to see if any parts will fail on it. A car is more important then a benching or overclocking computer.
Not to me it isn't. PC makes me money, car just sucks it up.
Cost of car = FREE
Cost of PC= $2k...
Importance of pc vs car...pc wins
Car dies...no loss of income.
Pc dies, guaranteed loss of income.
HMmm....wanna rephrase that?
Tested only 30 minutes of Prime95, because this time I can run any test bench and not have error in any core.Quote:
Originally Posted by laragirl83
This is stability for me, not beholden running days and days of Prime95 to prove nothing to nobody mainly to you.
Try it and post here.
Sorry for the bad english
here my e8400
http://img117.imageshack.us/img117/4...utosdq7.th.jpg
Quote:
Originally Posted by M@deira
If you only need 30 or 60 minutes of prime to suit your needs, because you personally have found that at that level you usually dont have problems, no one really cares. Not to mention if you do have problems, you can just bump up vcore then. Just like no one cares if someone else needs 8-12 hours of prime to suit their needs, because at that level, their personal use has shown that they dont have problems only then.
But, I think you need to read the context in which someone posts, before concluding someone is asking you to prove something. The original person was complaining of running prime 50 minutes once, than worried it was degrading because it only ran 3 minutes second time. And the RESPONSES TO THAT were that 50 minutes of prime is not long enough (given variability of short runs) to prove you are stable enough to repeat it and thus you can not conclude you are degrading, when it is more likely you simply did not have high enough vcore to consistently run prime stable over an hour. For instance, maybe you picked a voltage that would have run prime for 8 hours, and thus can consistently run over an hour. Or maybe you picked a voltage that would have stopped at 62 minutes, the next at 8 minutes, the next at 40 minutes, the next at 3 minutes.
Then you chimed in complaining that someone was asking you to prove something???
As an aside, why prime at all if only interested in stable enough to bench...I can bench 32m spi, 3dmark6, 3dmark5, pcmark5, at a point where I not even prime/orthos stable at 2 minutes.
My final overclock with a broken sensor.
http://i51.photobucket.com/albums/f3...an/4200mhz.jpg
Quote:
Originally Posted by laragirl83
Quote:
Originally Posted by cadaveca
Why rephrase? the major part of us here dont care if the cpu are stonerock superduper solid 24h+ , most uf us also do have a less benching computer aside from the monster bench machine most of us do got.
i can also do alot of stuff with less vcore than the prime can take.... the word "stable computer" are whats stable for YOU as a person, theres no rule for this, dont think u can call other peoples cpus "crap" just couse you think you want a PROOF of 18h prime, why do that? and if you are so addicted to your comp, why running it OC at all? risking it to crash if its so important.
well well, u cant come to XS and complain why people dont show of 20h of prime haha..
nice OC with that low of vcore, and nice VID...below is my orthos run at 1.32vcore and 4.2ghz...wanna trade?:DQuote:
Originally Posted by zizo
Quote:
Originally Posted by cadaveca
Not really, do you walk to work? Computer worth more then your car??
That's called $50 socks and no shoes.
zizo: 4202MHz @ 1.32 volts is excellent. Maybe you could add a picture to the Wikipedia and show the world what a grade 'A' E8400 looks like! :up: What's your date and batch code (Q7--A---) printed on the top of your processor or from your box? I've run my E8400 at 4200 MHz but to be Prime stable instead of needing 1.32 volts I'm probably going to need about 1.42 volts. I haven't decided whether I'd push it over 1.40 volts for 24/7 use. Probably not.
I've noticed during testing that whenever Orthos fails it always fails on Core0. That made me realize that Core0 is definitely the weakest link but I was curious to know just how weak is it?
I dropped the core voltage to 1.28 volts and loosened the timings to CL5 and booted up at 9x401MHz. I decided to run SuperPI on each core individually by doing a SetAffinity... in TaskManager and forcing SuperPI to run first on Core0 for some testing and then forcing it to run on just Core1.
Using SetFSB, I kept bumping up the MHz until SuperPI would fail on each core. Because Core0 seems so weak I thought this test would show a large difference in each core's abilities but it didn't.
Core0 could do a SuperPI 2M run repeatedly at a FSB of 449 MHz. As soon as I bumped it to 451 MHz, it failed.
http://img218.imageshack.us/img218/3...1mhzmz1.th.png
I tried the same thing on Core1. It could run at 451 MHz and 453 MHz but at 455 MHz it failed and locked up the computer.
http://img90.imageshack.us/img90/384...3mhzad5.th.png
This test confirmed that Core0 is the weakest link. The FSB where each core could run SuperPI 2M reliably at 1.28 volts was only a difference of 4 MHz (449 vs. 453). Not really a good enough reason to EBay it off. You might want to try this same sort of test using the single threaded version of Prime95 and then swap it back and forth between cores until you find your weakest core. SuperPI benches like this would also give you a good baseline before priming at high volts and high MHz for 82 hours. It would be a great way to test for a degrading CPU, sometime in the future.
To finish off testing I reset everything to my 24/7 OC and did a couple of SuperPI benches. I'm hoping some other users can do me a favor and post their 1M and 8M times at 9x450 using similar memory timings or even with DDR3. I think my P5B Dlx is still competitive and I'm interested in finding out how it compares to some of the newer Intel boards when the clocks are more or less equal.
1M = 11.515s
http://img231.imageshack.us/img231/4...hz1mbe5.th.png
8M=2m 28.937s
http://img149.imageshack.us/img149/8...hz8mwb3.th.png
Just got my E8400 two days ago. Trying to find the max FSB without increasing the vCore one bit (set at 1.1v in BIOS). So far my results are 6x500, but I haven't stability tested it yet (hoping to find SS max FSB first and then work down). System is in the sig.
Ohh, also first OC related post :)
What is Wr on E8400 and air ???
i was enable to break 5107MHZ on asus commando and i was near even 568 x 9 to but when i save it so i got black screen, motherboard and vcore at 1,7v reading in bios, but when i forgot to update my cpu-z to ver 1.44.1 then was my cpu.cvf file was not uploaded correct at my validate :( so i must try again and take it
True, for me as well stability and reliability of my machine is above all. I don't want a random crash EVER (been there, done that). So I'm willing to sacrifice a few percentage points of performance for stabilityQuote:
Originally Posted by cadaveca
A while ago people were posting about changing VID's, has anyone found out what is going on with this???
How much voltage r u applying on the NB and other chipset stuff.Quote:
Originally Posted by Ajarn
Auto's. I'll go check them though after my RAID is done being verified.Quote:
Originally Posted by RedBull78
My times at 9x450 and memory at 5-5-4-12 are:Quote:
Originally Posted by unclewebb
1M = 11.312s
8M = 2m 37.016s
Just obtained these results now, I didn't do anything while they were running, but I've got all my normal apps open.
Completely agree, that's why I haven't gone much above 4ghz, that's where I wanted to be and stability was more important to me then another 5-10% in mhz.Quote:
Originally Posted by user1234
Few things more irritating then having random problems with you computer, won't come out of power save after hours of sitting. Been burned before chasing what I thought was drivers / software only to find that the extra 10% put me on the edge with heat and just the room getting hotter tips the apple cart.
Checked and these are my voltage setting in the BIOS:Quote:
Originally Posted by RedBull78
SB 1.05V: 1.070v
SB Core/CPU PLL: 1.55v
NB Core: 1.33v
CPU VTT: 1.100v (CPU vCore: 1.13v)
Clockgen: 3.45v
Got my e8400 yesterday and have been having a blast playing with this little piece of silicon :D
Batch# Q746A488 and its a good one!
http://i37.photobucket.com/albums/e82/sonofander/45.jpg
now before the degredation police show up and give me what for for my vcore... :p:
seriously, this chip only has to last me a week before my 9650 gets here, so I'm just having fun... I figure under 1.5v for 24/7 will get this chip through the next week right??? :rofl:
I'll let you guys know how it goes
yeah but might as well save the life and sell it...
or perhaps I can single handedly dispell or prove a much speculated idea that these chips are falling apart in our little overclocking handsQuote:
Originally Posted by DamienKC
Imho, your chip will be just fine for a long time to come, but I don't think that'll change anyone's mind. Nice overclock :up:Quote:
Originally Posted by sonofander
that would be my guess as well...Quote:
Originally Posted by PhilDoc
Thanks man :up:
yea nice, let us know how it goes. tortured mine a full week around 1.425 - 1.45v and its still kickin a**s.. please post back your settings when successful. i personally think these badboys can do 24/7 @ those volts. i want 4.5ghz. nothing less. voltage isnt an issue for me. this is an experiment and my chip dying is just part of the experiments running cost lol. we are overclockers. where is the fun in playing safe lol. am new @ xtreme and was very disappointed to see how un xtreme most users are. talk about irony. i wish someone could start a new thread saying extreme Ocing of these chips.
Looks good. What are your settings in bios for that overclock?Quote:
Originally Posted by sonofander
NB @ 1.57
PLL @ 1.5
FSB @ 1.3
vcore @ 1.5 (1.48 real)
all other volts @ auto
Not sure what else to tell you...
I finally got around to doing my 24hr. stability run! Keep in mind that this is what I have been running this chip at for a few weeks now, under water without any issues. My chip is definitaly not having any kind of issue with the voltage degrading my OC. This is at 1.45v and no vdroop under load, measurement was taken with a DMM.
http://img257.imageshack.us/img257/6...4637sn1.th.png
Nope, I drive about half the time. I got an 84 Buick Cutlass w/442, guzzles the gas. City I live in has population of just over 1 million...transit sytem is pretty good.Quote:
Originally Posted by dogsx2
When my best man's father died, I got the car.:( Long story, but the original onwer lives on through the car, so it's not a daily driver.
Sure, the car is WORTH more, but it COST me nothing, other than the $800 a year for insurance. I'll spend that on my pc every two months!
I don't use credit, either, so buying a new car just doesn't appeal to me...insurance would cost more!
Anyway, I can't have my pc crash doing a render...some take many hours, so I prime for that long. If it passes prime, It'll render without issues, at least so far.
Call it what you will, it's obvious you and I aren't gonna agree, so I'll not waste any breath and discuss this with you further. I do alot of things backwards from some people...it's 10:00pm, and I'm about to go to work...Have a good one!
just tried a quick voltage bump out of boredom and booted this no problem. seriously these chips rock.
http://valid.x86-secret.com/show_oc.php?id=314803
Has anyone found out how to fix this temperature stuck on the two cores yet? Right now, my Q746A483 has a stuck temperature of 42 for core1 and 40 for core2. CPU temp is way below motherboard temp, which I know is false. If I run torture test on them, they will go up but they never go below the temp above even when they are idle.
Where do you guys get CPUZ 1.44.1? I only see 1.44.
We can do that when our RMA'd IP35 Pro shows up - here's our Commando @ 5-5-5-15 on Vista 64 :)Quote:
Originally Posted by unclewebb
1M = 11.607s
http://img115.imageshack.us/img115/6...0mhzkn2.th.jpg
8M=2m 36.702s
http://img213.imageshack.us/img213/3...0mhzuv6.th.jpg
http://www.cpuid.com/download/cpuz_144.zipQuote:
Originally Posted by jcniest5
google is your friend :)
I wished I had exchanged my Q748A223 for one like yours (which I saw two of them). Instead, I did it for a Q746A483 and I couldn't even boot at that setting. I even tried a suicidal Vcore of 1.7V and the system can't even boot to Windows. How very unlucky I must have been with these technie thing!!! :mad: :mad: :mad:Quote:
Originally Posted by DamienKC
Well, it seems that 510mhz is the highest FSB I can archive without touching the vCore (unless I start messing with some more chipset voltages). Only voltage changed was CPU VTT, and that was to 1.267. Now to see max FSB with touching vcore ;)
Well, found the highest FSB (though probably not stable, just enough to post this) that I'm willing to push to.
BIOS Settings:
VID Special Add: 250mV (Translates to 1.44v in BIOS/1.456 in CPU-Z)
SB Core/CPU PLL: 1.55v
NB Core: 1.50v
CPU VTT: 1.600v
Clockgen: 3.45v
Final FSB: 525mhz
nice! but what will that voltage get u for core speed ?
Nice VTT!:DQuote:
Originally Posted by Ajarn
wow 1.600vtt is way to much man, when I tested mine i could run with 1.20VVT set in the bios on the maximus which is 1.13 ( it droops) and could run up to 500fsb stable.
i Ask before what is Wr on a e8400 with air ??? i dont have seen a e8400 over 5,1ghz yet
My current 24/7 speed for my Wolfdale is 4600mhz at 1.47v on its own massive water loop. I can bench at 4800 at 1.625v I just finished a run of Orthos small FFT at 4510mhz which needed 1.50v but could run all day long and my temps never even got into the 50s though this Computer room is rather chilly. I have been running this Wolf at 4500+ now for about a month with no signs of electro migration break down yet. Really incredible little chip for the money.:D
I have 1M time recorded for 9x450 at 24/7 timings/settings that I use for 4.1 (and for 4.05), done with all normal startup apps running. 3 runs 11.422, 11.422, 11.421.Quote:
Originally Posted by unclewebb
I just built two identical E8400 + Gigabyte DS3R systems. Everything is the same, with the heatsink being the only difference -- one has a Thermalright Ultra 120 Extreme and the other has an Ultima 90.
(1) The Ultima 90 system has a CPU Vcore of 1.18v. Both cores are reported at 23-25C (idle). It runs 3.6GHz with stock 1.18v @ 28-30C (idle). All power saving features are disabled.
(1) The Ultra 120 Extreme system has a CPU Vcore of 1.12v. Core #0 is 38c while idle and core #1 is 25C. It runs 3.6Ghz with 1.20v; core #0 stays at 38c and core #1 increases to 35C (idle). All power savings features are disabled.
Is such variation normal between two chips from the same bin? Or did I not install my Ultra 120 Extreme correctly?
has anyone tried Q746A483 yet?
any comments on it?
Nice 24/7 OC...and thanks for info...only way we are going to find out if its just some chips or all chips sensitive to moderate voltages, is if some are still running them in that range and reporting info.Quote:
Originally Posted by flesheatinvirus
Given that one chip became unstable after 3 weeks of running only all auto/stock settings, may be a quality issue in few chips that is contributing, and maybe that issue will improve over time. Though with ~8 chips degrading running long orthos runs ~ low 1.5's, I am thinking 1.5+ is not going to fly for most chips at 24/7.
nice chip!Quote:
Originally Posted by flesheatinvirus
@ 1.5v I can only stabalize at 4.55 :(
trying like hell to get 4.6 going here, but as crazy as I am, I'm not willing to go past 1.5v for daily use
I agree unless you have at least hard core water or better 1.5+ will be an issue for most longterm. My Wolfy is a Q745A576 1/4/08 pack and seems to like 4500-4600mhz at 1.45-1.46v, beyond that requires a much larger increase in juice which is simply not worth the extra heat except for an occassional speed run. Chip settings are as follows. CPU PLL V- 2.20, FSB Term V- 1.40, NB V-1.55, SB V-1.20, Clk Overcharge V-1.0 and I run Load line Cali disabled which seems to work better then enabled with this Wolf compared to my Quad. CPU and NB GTL Ref set to minimums of 0.57 and 0.61 which certainly helps stability on this P5K Prem beyond 500fsb, max so far is about 540 though I have not tried a lower multi combo yet.:D
2.2Vpll? That's gonna kill it slowly mate.
don't get the Q746A446 01/07/2008 :down:Quote:
Originally Posted by stephen23
Good point 2.2 is too much, I dropped it to 1.9 right now for another test and will see just how low I can get it.:)Quote:
Originally Posted by Zeus
what are your setings like? mine needed 1.5v actual to prime stable. hopping to lower it.Quote:
Originally Posted by flesheatinvirus
My settings are in my other post above my last one though CPU PLL is at 1.90v.
Ken_F wrote:
Same thing with mine, Ken, it's probably just a fault with a core diode not reading the low idle temps well. My first E8400 had core readings of 17C/34C at idle. I exchanged it for another one from the exact same batch and it's cores read 38C/38C at idle @ 4.0GHz... (but one core sometimes flickers down to 32C). But the new cpu with the higher idle temps shows lower load core temps of 50C, while the old cpu got to 60C under load with the exact same settings? I just ignore temps now and don't exceed the 1.365v recommended by Intel @ 4.005MHz and all is well.Quote:
Is such variation normal between two chips from the same bin? Or did I not install my Ultra 120 Extreme correctly?
With regards to VTT (FSB Term), isn't the maximum voltage 1.115v? I have mine set to 1.1v and my E8400 o/c's to 4.0GHZ stable as a rock. Putting 1.4v or 1.6v through these tiny circuits is cruel and unusual punishment!
can someone post all the most important settings in bios for ASUS P5K-E...what each setting is for.....maybe some pic too?
Since when are cars free - unless you're living with your parents? Of course, if you're still living at home, I doubt that your PC is making you THAT much money :pQuote:
Originally Posted by cadaveca
The guy having funny temps between two e8400's - some chips may seem cooler. But I have a feeling it's not reporting the right temps.
My :banana::banana::banana::banana:ty 1st chip took 1.36Vcore - never hit 4Ghz properly (3.9xx), and load was recorded as 47/51 in OCCT. Sorry but that's bs haha. My 2nd chip is at 1.29Vcore - hits 4.005Ghz, and loads at 62/61. I think you can understand which temps I would trust more...and it ain't the first.
Ditto. I actually have mine at 1.15V, but I'm sure even that's not even needed. Anyone who's OCed properly (isolate the variables) will notice that these chips will scale fairly well with just Vcore. Maybe add to NB if you're going FSB heavy...but that's dependent on your mobo and not the chip itself.Quote:
Originally Posted by Xvys
With 1.2VTT I could get 510 easily. Was just going for a SS ;) It's back down to 1.100 for finding max mhz.Quote:
Originally Posted by mc2k
Yes, there seems to be a fairly large variance between different E8400's, mine idles at ~35 at idle on stock, ~40 at 4.2Ghz..Quote:
Originally Posted by Ken_F
Very nice speeds and temps. What is your ambient temp?Quote:
Originally Posted by flesheatinvirus
A short run under SS .
E8500 was a Q740. I think it's the worst batch possible for SS. 5100mhz is lame for -50C :( Very good score at that freq. non the less :)
Setup:
http://ocidb.com/albums/userpics/101...l_IMG_0007.JPG
Out of curiosity being a Commando owner, what max fsb can you boot & how much extra fsb can you dial in on set fsb?Quote:
Originally Posted by bonzo_catalin
I see your link width is @ x1 too is it luck of the draw with the DFI as to whether the width stays up ? I ask as my Commando drops to x1 sometimes too.
CN :)
We only booted at 475fsb and used setfsb from there. Didn't try more... I think it was showing x1 width because the vid card was in the middle PCIx slot, not 16x one.Quote:
Originally Posted by C-N
My Comp room is in the high 50s F which is pretty cool.:)Quote:
Originally Posted by g1raffe
Moved onto finding max mhz under a SS (without SetFSB, I want boot max). So far I got to 4Ghz with these voltage settings:
vCore: 1.34 in BIOS 1.360 CPU-Z
CPU VTT: 1.233
NB Core: 1.50
Rest same as before.
I don't want to push the vCore much above where it is, because I plan on running my OC 24/7 and getting a high mhz with an insane vCore does me no good in that respect.
Edit: Well, it seems to be at least decently stable, an hour of Orthos so far.
http://img408.imageshack.us/img408/2...ableyx8.th.jpg
Here's a bit I found on Anandtech about an E8400 and some different VTT and GTL scenarios they used. But yeah I saw 1.6 and was like :eek: .Quote:
Originally Posted by mc2k
Ajarn... why run an 8x multi? Doesnt this stree the FSB too much? I am new to OC and what to learn more like when to change multi.
What do you have the memory set to? linked/unlined? what ratio?
lower multi = higher fsb for the same clock = higher bandwidth. good overall performance specially memory
Ajarn very nice 4.0 without going past 1.360 volts ! hope mine comes in at least that good
Since I tested the FSB up to 525 last night I though it would be fine to get to at least 500. And yes, with a 500 FSB I'm getting pretty good bandwidth :)Quote:
Originally Posted by Hughyhunter
Ratio is 1:1 (the memory can hit 562 with the current voltage).
Thanks! Yeah, I'm glad that I got it to at least 4.0. If I get anything more with a higher multi or whatever it will just be icing on the cake :DQuote:
Originally Posted by HDCHOPPER
A quick test
Vcore 1.4 - VTT 1.13v - 4635mhz - 9.9s superpi 1m
http://img338.imageshack.us/img338/8562/1moi0.th.jpg
I'm currently running NB at 1.5v, just wondering, is that safe for 24/7 without active cooling? What should be the max? (I see a lot of people on the net use 1.61v as in the Anandtech article, but that number looks a bit large to me)
Edit: Actually, comments on any of my voltages would be nice ;)
eyeve got 1.61 on my n/b for 24/7 but it does have a fan... eye put a fan on everything ! s/b...n/b (mars has a fan on n/b allready but eye put on another for 2 ;) ..hard drive 2.... memory fans ..3 extra lil cpu fans on the backside of vid card hehehehe volt reg area evertyhing ... wouldnt be overclocking without fans
Q743A748 01/04/08 from NCIX on the first day of release.Quote:
Originally Posted by unclewebb
Well seems that for 4.0 that 1.5v NB and 1.233VTT was unnecessary, got away with 1.200VTT and 1.42v NB. 1.5 Hours OCCT Stable. (Temp was 39C from delta, forgot to get a screen just before the cpu cooled down)
Excellent sub 10 superpi!Quote:
Originally Posted by mrcape
Could you post your bios voltages and memory timing? I have a nearly identical setup. Also what bios are you using?
THX! 01/09 biosQuote:
Originally Posted by scifikg
This clock could be tuned to drop some of the voltages and still get sub 10sec. Or maybe up a couple fsb notches without upping Vcore. I just went for this in between 3dmark tests. I was trying to do some quick stuff with vcore around 1.4v max. I gotta get this board and chip really cold and take some benchmarks soon.
Here are the bios set voltages you requested for the superpi cap above -
Vcore 1.4v
VDIMM - 2.31v (could lower this)
SB - 1.07v
PLL - 1.55v
NB - 1.45v (could lower this)
VTT - 1.15v
And mem for that pi run was 5-5-5-15 1:1 1090mhz. Anything below FSB 540 I was running 4-4-4-12, this GSkill is badass, I need more.
Good luck!
BTW, you may find these other settings useful -
Bios Screens for my 24/7 - 4.25ghz @ 1.28vcore
http://img518.imageshack.us/img518/7...425gzk1.th.jpg
Ajarn:
Is there a reason your CPUID shows 4Ghz but OCCT shows a more startling 4.5Ghz?
Always trust CPU-Z. OCCT assumes he is using the default 9X multi for the E8400 but CPU-Z is smart enough to know he is only using an 8X multi.
Didn't read all the thread, did ya.:stick: The previous owner died, and left it to me.:rolleyes:Quote:
Originally Posted by dosbox
And I explicitly stated cost, not worth.
And I've got 4 kids of my own, that I support, plus my wife, on my meager wage and money made from my pc. I charge $25/hour for renders for local artist...which is more than I make/hour at work, actually, so yeah, pc is far more important than my car. Drove yesterday, took the bus today...
Anyway, since you are so quick to judge, without first reading all the info, you've been added to my ignore list. Have a good one!
More to the thread's topic, my first E8400 is now D-E-A-D. no post, period. 2nd(L746) is churning away @ 4ghz,(500x8) 1.375v. Temps are within 2c on each core, 54c loaded under DTek Fusion block, MCP655, and Swiftech 2x120mm rad.
Seems they are starting to die. You benched it at high volts right?Quote:
Originally Posted by cadaveca
not until after I noticed it degrade. Maximum given was 1.55v, it died shortly after...4 days maybe.
It was priming when it died tho...not even BSOD, just black screen, fans still running..blah blah...
Noticed degredation @ 1.425-1.4375v, tho. It was a bad cpu from the start...and it has in no way stopped me from clocking my other 2...I think it was jsut a bunk chip that fell through the binning process. Worked fine @ stock when i got it...
Chances you take when overclocking, period...cost me $200...oh well.:shrug:
So, is 1.4v still considered the max "safe" vcore for the 45nm chips? Has anyone's died/degraded yet from 1.45v+ ?
@cadaveca, just noticed your post - Perhaps I should leave my vcore at 1.39v (lowest I can get for 4.1ghz on air)
i edited, sry, maxwolfie. To me, seems like 1.5v under really good water is pushing it...i'll stay under 1.45v with water, 1.375 on air.
Possibly...the death was caused by the board, not the cpu itself...pls take note that INtel has already stated they had issues with FSB scaling and cheaper boards...don't know exactly what that problem was...
Further to your post, it seems as though temperatures really aren't a problem. Regardless of cooling (air/water/whatever), I still wouldn't push it above 1.45v volts, as per Intel.Quote:
Originally Posted by cadaveca
I reckon your degradation would still have occurred even under some form of extreme cooling. What do you think?
<just dropped multi down to 8, orthosing at 4.1ghz atm, 1.39v, so far so good>
EDIT: 4.1 stable @ 8, 4.2 not. Might as well go back to 9x
hard to say. I don't fully understand WHAT killed the cpu...voltage played a role, sure...but why for my cpu, and not others?
I'm hoping maybe one of my others will start to show signs of death, and then I can maybe pinpoint the problem, but i really think this is far out of my scope of know-how.
Extremecooling hasn't prevented death of some QX9650's...yet others ahve been given over 1.8v, without reported issues...
I don't think there's any real way of testing it. All cpu's are different in some respect.Quote:
Originally Posted by cadaveca
Hey, it's what makes overclocking interesting! :)
Everyone seems to have different results when clocking these chips and there has been a few deaths so far though I have not experienced any performance loss over a period of a month which is good considering how high I have clocked it. Im currently testing 4302mhz at 1.38 bios which so far is bench solid, Cinebench R10 just did 9142 and 3DM06 with stock 8800 was 14,243. Orthos is up next.:)