**Official e8400/e8500 Retail OC Thread

[Viss]

yes its weird i have tried 3 new e series cpus and they all seem to go to 4600 but no more some with lower voltage but after 510fsb they get unstable no matter what volts on nb or cpu

Think its board/bios. Tried 3 that all can do over 580fsb on commando but nothing much over 500 on maximus. With me E6600 and bios 0401 maximus did 550mhz air and 590 on dice (for 1M) but with wolf and never bios even 500fsb is hard. I might give bios 0401 a go again to see if it helps but i want the half multi

[GLENBOY]

thanks loco , i thought it was me or just a run of bad luck, i have the asus max extreme and a few others who have had them have found over 500fsb unstable, i am talking on the max multilplier, i might try an older bios also maybe or wait for a newer one,still not bad thoiugh but the p35 boards seem to have it a lot easier with these at higher fsb

[giorgos th.]

With any maximus you`ll get fsb issues...
i got 527 max on my formula @ 0907 bios.

[Spyrus]

Confirmed but only for Wolfies.

[NINaudio]

I'm running an E8400, batch 748, at the speeds listed in my sig.

cpu temps are 40/56 idle/load.
NB temps are 30/36 idle/load.

W/ NB @ 1.55v I was able to get an hour of Orthos at 4.0 using 1.39v, and an hour at 4.1 using 1.43v.

You guys think a bump in the NB voltage to 1.7v will get me a higher fsb without more vcore?

[Spyrus]

In order to exclude nb voltage out of question lower tRD (performance level) to 8.
To do that set Relax Level to next +1 value.
Example in my setup:
9x490@strap333@nb 1.51V@disabled/0
9x490@strap333@nb 1.33V@disabled/1

[rge]

After running numerous orthos runs on my E8400 trying to find stable settings, imo, it is difficult to discern variability from degradation if relying on orthos runs less than 12 hours. While I have no doubt in lieu of some dead chips, that degradation is occurring in some, I am wondering if some of the reports at least with more moderate voltages are seeing variability, like I saw, and not degradation. Examples:

Testing 4.05, 1.31875v bios
Run 1 = orthos shut self off at 10 hours (logs).
Run 2 = orthos fatal error message at 3 hours.
Run 3 = orthos system rebooted 11 hours.
1.3325v bios got same variability in runs.
1.3375v bios 2 runs at 14hrs+ stable

Testing 4.1,
below 1.3375v get stop messages
1.3375v bios orthos system reboot 10 hrs 50 mins
1.34375v bios orthos orthos shut down 6 hrs
1.35625v bios system reboot at ~ 5 hours, second run 9 hours then reboot while surfing
1.3625v bios 1 run stable 14hr+, second run commencing.

I am not saying you need 14 hrs of orthos to get a stable enough computer to run games, general usage etc. In fact, with this peculiar chip, above 4 ghz there may be a range of volts at given mhz where the machine is perfectly usable, but just cant get repeatable orthos runs.

My point is, some reports of degradation (not all) may be the same bizarre variability that I witnessed using orthos runs shorter than 12 hours at the voltage range (at given mhz) where you stop getting stop message and instead getting shutdowns, etc, which seen to occur at random times, instead of increasing long runs as voltage increases. And I am wondering, if to prove degradation, runs of 12 hours or more are needed. (dead chips and things like use to boot at 1.2v, now wont boot at 1.25, aside)

I am not expecting degradation at the VID limit, which will be my 24/7 below, but I will check in time and see.

[dissident]

well if you're looking to sell your chip for something else, they have great resale value on ebay.. mine lapped sold for more then all the buyitnow auctions. I've noticed this with other auctions too where the auctions close higher then what you can buy them now for.. weird.

[Chri$ch]

Confirmed but only for Wolfies.

yes, but it isnt only the maximus... much boards have problems with high fsb if you
use a wolfdale.

My P5K-E is the first P35 Board what can do more than 500FSB stable with my E8200,
all other i tested maxed out @ 500FSB (DFI LP DK, Asus P5K Pro, Asus Commado,
Maximus Formula (SE and non SE), Foxconn Mars.. etc..

[Spyrus]

After running numerous orthos runs on my E8400 trying to find stable settings, imo, it is difficult to discern variability from degradation if relying on orthos runs less than 12 hours. While I have no doubt in lieu of some dead chips, that degradation is occurring in some, I am wondering if some of the reports at least with more moderate voltages are seeing variability, like I saw, and not degradation. Examples:

Good find there i have same thoughts myself.
If the coretemp of yours is saying the truth (65oC) this is one thing to note.
If you can lower temps and use the lower vcore used for prime stable 2+ hours
maybe you can solve the problem. E6x50 series could do prime at around 70-oC
but Wolfies seems to be more sensitive to temperatures.
If temperature is not your problem then i cannot think something right now.

Degration though maybe can be reported easily with lowest vcore a system can boot (like you noted and
i agree on that).
If someone writes down minimum values and have after days degration signs then
he could try to boot with known lowest vcore.

[Infa]

Dude, that was a general statement. Not particularly directed at you. Anyway, enjoy!

Ok, sry man...

Im a react man, i take stuff to personal sometimes...
hehe..my bad.

regards.

[NINaudio]

In order to exclude nb voltage out of question lower tRD (performance level) to 8.
To do that set Relax Level to next +1 value.
Example in my setup:
9x490@strap333@nb 1.51V@disabled/0
9x490@strap333@nb 1.33V@disabled/1

Thanks Spyrus, I'll give this a shot when I get home.

Were you running your cpu at the same voltages as well?

[unclewebb]

rge: Thanks for thoroughly testing and sharing your Orthos findings and putting another nail in the degradation myth. I know with the previous generation that when I started to approach the max my E6400 was capable of I really had to watch temps to maintain Orthos stability. When air cooled at 3200 MHz I could run Orthos stable at 80C but at 3400 MHz I was down to 70C max. When running at 3600 MHz I'd have to keep the core temps under 60C to maintain long term Orthos stability. When you're at the edge like this it's easy to be stable one day, then the sun comes out and the room temp changes by a couple of degrees and all of a sudden you're crashing and getting ready to RMA your degraded CPU.

Has anyone had any serious degradation where they can't even run 4 GHz anymore that is not related to a memory module going bad or some other problem? Hopefully I'll finally have some time this weekend to get my new E8400 Q746A500 into service and give it the voltage and MHz that these things are crying out for! Weekend plans include some naked testing of the core to see whether CoreTemp is as accurate as it was with my E6400.

[cadaveca]

You work for Intel, or what? Seems like you are on a damage control campaign...

If you can prove that my cpu has not degraded, I'd be more than happy to try what you request...I've now gone up to 1.45 for 4GHZ to be stable...watercooled, of course.

[Zeus]

You work for Intel, or what? Seems like you are on a damage control campaign...

If you can prove that my cpu has not degraded, I'd be more than happy to try what you request...I've now gone up to 1.45 for 4GHZ to be stable...watercooled, of course.

1.45V? That's an awefull lot. Coming from?

[cadaveca]

From a wee bit less than 1.35v. That's a .075v+ increase in needed voltage over the space of about 1 week.

Don't get me wrong tho..I think only a few chips will degrade like mine has...I've got the largest loaded temp difference reported between cores than anyone else so far...15c-17c...as well as the highest "stuck" temp.


But in the last month or so I've heard of more QX9650's dying than I have EVER heard of QX6700, or QX6850 dying. Coincidence? Maybe, but this is coming from guys that know how to properly treat hardware...I bet almsot every single guy in top 10 has killed at least 1 QX9650, while maby 3 killed 65nm chips.

[Spyrus]

Thanks Spyrus, I'll give this a shot when I get home.

Were you running your cpu at the same voltages as well?

Yes the only thing changed was PL and nb voltage.
Everything else no touching.

[Spyrus]

From a wee bit less than 1.35v. That's a .075v+ increase in needed voltage over the space of about 1 week.

Don't get me wrong tho..I think only a few chips will degrade like mine has...I've got the largest loaded temp difference reported between cores than anyone else so far...15c-17c...as well as the highest "stuck" temp.


But in the last month or so I've heard of more QX9650's dying than I have EVER heard of QX6700, or QX6850 dying. Coincidence? Maybe, but this is coming from guys that know how to properly treat hardware...I bet almsot every single guy in top 10 has killed at least 1 QX9650, while maby 3 killed 65nm chips.

Did you record you minimum booting voltage before and after degration?
I am interesting to see that is the same difference with voltage needed to get stable.

Example:
3.6GHz@1.1 Volts minimum voltage to boot
After degradation:
3.6GHz@1.175 Volts minimum voltage to boot

[unclewebb]

I'm not saying that CPUs don't degrade. In my experience though, over volted D9 memory modules seem to fail a lot quicker and more often than over volted CPUs.

rge's numbers show the huge amount of variability in hours of Orthos stable. I hear too many people jumping to conclusions based on some rather short and very unscientific tests. rge went 10 hours Orthos stable and then only 3 hours stable during the next run. Too many people would be screaming "it's degrading" after something like this but then on his next run he went 11 hours.

It's possible that at 1.40 volts and beyond these CPUs will continue to degrade like the Intel docs seem to hint at but it's also possible that it might reach a point where it levels off. Maybe with 1.45 volts your processor will run at 4 GHz for years to come.

If your processor degrades some more you can send it to me and I'll be happy to test it for you. Edmonton is only a few hours away so maybe some day I could bring up my computer and we could install your CPU in it to get a second opinion of its health status.

[cadaveca]

Unclewebb...LoL. I;ve got literally $20,000 worth of cpus, memory, boards, vidcards from the last year sitting here...I'm not the sort to just rampantly post over an issue like this without verifying the info. I post when I find truth, so that others may know, or to incite conversation...but with something this serious...that may put some users off of making a purchase of some of the fastest silicon there is on the market...that's just not me. I want more records broken, not stunted sales of good parts. The only allegiance I hold is to performance.

Spyrus, inital 3.6ghz boot voltage(not stable) was 1.1175v. It now requires 1.225 for boot, and 1.275 for stability. So @ 3.6ghz the degredation is far greater. It really makes no sense to me...other than I got a bunk chip. But that happens...in fact, I'm off to buy 2 more e8400's as soon as the local store opens.

I've tested cpu in maximus formula, and MSI p35 Neo2FiR, as well as P5K vanilla. same voltage required on all three boards.

I've also lapped my cpu(to guarantee no RMA on this one), and noticed a 5c drop in load temps @ 1.4v(old 4ghz stable), however not one single MHZ more in clocks from that drop i ntemps. Like 4ghz is stable, but 4050 is not..altohugh when I first got the cpu, 4050 WAS stable at the same voltage.



I mean no offense by my posts unclewebb, however, you are posting on this issue far too much for someone who hasn't even stuck a 45nm cpu in a board yet. Persoanlly, any posts you have made on the issue are not truth...you have NO experience with this chip, or if you do, you've been hiding it.

EDIT:
BTW, you HAVE been saying it doesn't happen...in fact, you jsut a few posts above called it a myth, which, unfortunately, it isn't. i WISH I could still even just bench @ 4.5ghz...as you can see by my sig, I was 3dmark stable, at least...now it won't even boot!

[Spyrus]

I think orthos is degrading Wolfies (joking)

[mrcape]

After running numerous orthos runs on my E8400 trying to find stable settings, imo, it is difficult to discern variability from degradation if relying on orthos runs less than 12 hours. While I have no doubt in lieu of some dead chips, that degradation is occurring in some, I am wondering if some of the reports at least with more moderate voltages are seeing variability, like I saw, and not degradation.

Just to get a handle on the accuracy of your data, did you say you were surfing during one of these tests? I think it would be best to set each run at priority 9 and surf on another machine while the tests run, ensuring the same conditions for each test.

[cadaveca]

TBH, i don't think it's voltage...it's current some cpu's will not draw teh same as mine does @ same speed..i think these are less likely to degrade. I also think that cpu's with sensors reading approximate same temps will degrade less as well.

however, it's gonna take quite some time before there can be any "medium" by which we might be able to tell if a cpu will degrade or not.

It's also possible that bios functions, set for 65nm current consumption, and not 45nm, is causing degredation...my next couple of cpu's should either confirm or deny this...if even one of them degrades, and fits the criteria I mentioned above for not degrading, I'll happily blame the board.

mrcape, you reading my mind or what? . maybe the spikes in PWM from such usage are an issue...

[ante_ante]

E8500 @ Q746A363

Testing some orthos under Vista.

Vcore 1,4250
455x9,5

[deathman20]

Hmmm curious about this degrading issue.

While I've basically had mine at 4.05Ghz with 1.37V Bios (1.21V windows load) since I got it 2 weeks ago. No issues with it so far. Haven't re-tested orthos stability but I know my temps full load is around 50C and idle is just below 42C range (can't tell since thats where core temp is the lowest reading it will give for me).