E8400/8500 degradation myth possibly busted?

[dnottis]

I havent read the entire thread... But I'm very concerned of this matter, as Im studying and cant afford equipment all the time - and lets just say, I've really really put an effort into making contact with Intel, but without success...

To my knowledge, based on information from OCXtreme, this started back with the 65nm C2D / C2Q, but only at very high voltages... Kevin, OPB (onepagebook) has tested this, and he says its due vPLL, not vCore... Cant remember his figures with 65nm, but 45nm, the one of importance for us - its not recommended to go past 1.7vPLL for dualcores, and 1.6vPLL for quadcores, if one doesnt do this, the vCore shouldnt have any impact...

- Before some of you get suspicous, no, I have not tested this myself, I havent even configurated a 45nm setup yet... 45nm was released very late in Denmark, I only got one a few weeks before Cebit, and directly after Cebit, I went for a marketing competition in France, returned this saturday,... Now my E8400 is up for sale, and Im going for Xeon....
- I intend to give it as high vCore as it will scale with on my watercooling setup....
For benchmarks, I've only done 1.8v - mobo didnt allow higher FSB for us to benefit from higher vCore...

What exactly is the CPU PLL voltage??

[aamar]

I'm wondering if the degradation is due to exessive priming at high volts rather than just high v with normal usage. Because lets face it, no application we use can consistently stress our CPUs more than priming for long hours.

[dnottis]

Just as a follow up, after having some issues after 1.50v. Changing a few things that were on auto, Im back to where I was.

[tranceaddict]

"dnottis" which things did you change, because I have the same board with the same BIOS and the same kind of issues as you. Thx

[CrazyNutz]

What exactly is the CPU PLL voltage??

"Phase Locked Loop" It's the Frequency signal connected to the CPU fsb.
Example: If you have the FSB Freq set to 400mhz the PLL signal will alternate
between low (~0v) and high (~1.5v) 400 million times per second. PLL voltage
is the maximum voltage of the "high state" so if you have it set to 1.6v it
will alternate between 0v (low) and 1.6v (high). Generally you would up this
signal voltage to make it less noisy (make the signal look cleaner and
stronger to the CPU, like tuning in a strong radio station).

However, I have found this to work just fine at the default 1.5v.

Also just FYI some defaults for the 45nm cpus are:
FSB Term Voltage: 1.10v (65nm was 1.20v)
PLL Voltage: 1.50v

I personally would stay as close to these defaults as possible.

[dnottis]

"dnottis" which things did you change, because I have the same board with the same BIOS and the same kind of issues as you. Thx

Well I think the main thing was the transaction booster, I had it on auto. I disabled it. The PLL was on auto too, now it's set to the lowest setting to avoid burning my chip up. I had some suggestions like just killing power to allow everything to completely lose power and clear itself. Check my lenghty post a page or two back. I've never had true degredation and was just stating some things I've had to do to get back to where I was when this "degredation" has affected me.

[mrcape]

The thing with PLL and FSB Term is that some bioses (designed for 65nm) lowest settings, are at or higher than the max intel specs. So users who don't know what those voltages do are likely to say "hey why don't I give all of these a bump" when they realize they are set at the lowest settings. Sad thing is, those "low" settings are actually medium to high.

[dnottis]

The thing with PLL and FSB Term is that some bioses (designed for 65nm) lowest settings, are at or higher than the max intel specs. So users who don't know what those voltages do are likely to say "hey why don't I give all of these a bump" when they realize they are set at the lowest settings. Sad thing is, those "low" settings are actually medium to high.

I actually forced PLL to 1.6v also, with Transaction booster enabled and PLL to 1.6v I ran another 11 hour pass.

[mrcape]

That's good to know that it didn't do anything to yours. I had read the same thing at the OCX site about PLL possibly causing damage, but I don't think anyone knows exactly how much will do it. The Intel e8 spec says 1.55v max for PLL but there's probably still some head room over that.

My comment about the bioses was more in general. The fact that the PLL & VTT (fsb term) voltage ranges never changed from 65nm to 45nm seems weird. People are bound to assume that the bottom settings are low.

[dnottis]

That's good to know that it didn't do anything to yours. I had read the same thing at the OCX site about PLL possibly causing damage, but I don't think anyone knows exactly how much will do it. The Intel e8 spec says 1.55v max for PLL but there's probably still some head room over that.

My comment about the bioses was more in general. The fact that the PLL & VTT (fsb term) voltage ranges never changed from 65nm to 45nm seems weird. People are bound to assume that the bottom settings are low.

Well the max suggested is 1.7v PLL max for 45NM duals and 1.6v PLL for quads. Stock is 1.5, the next jump is 1.6v on my board. I gave mine 1.6v, the first jump from stock.

Here was the quote I found on OCforums from OPB -

OVP/OCP Feature on Mobo is more important these days.and...dont use pll > 1.6v on quad , 1.7v on dual. (45nm )


Actually on my P5k when I select those settings it shows the min and max for 45 and 65nm settings. The new bioses that supports 45nm bioses will detect and adjust accordingly.

[CrazyNutz]

That's good to know that it didn't do anything to yours. I had read the same thing at the OCX site about PLL possibly causing damage, but I don't think anyone knows exactly how much will do it. The Intel e8 spec says 1.55v max for PLL but there's probably still some head room over that.

My comment about the bioses was more in general. The fact that the PLL & VTT (fsb term) voltage ranges never changed from 65nm to 45nm seems weird. People are bound to assume that the bottom settings are low.

The PLL voltage for the 45nm did not change from the 65nm, but the
VTT did. For the 65nm VTT is 1.20v, and for 45nm VTT is 1.10v.

EDIT: Sry I just realize you were speaking from the bios perspective, However
on my P5K-e the bios does change the Vtt range from 1.20-1.60 (65nm) to 1.10-1.50 (45nm)


E8000 spec document says:

Vtt min: 1.045v typical: 1.10v max: 1.155v
PLL min: -5% typical: 1.50v max: +5%

There is also a section that states:

Absolute Maximum
Vcc 1.45v
Vtt 1.45v

[mrcape]

Interesting about P5K bios. Is that a recent change or has it been that way since 45nm launch in January?

The lowest PLL setting in the DFI bios is 1.55v. The VTT minimum I think is 1.10v or 1.12v. They added new 45nm GTL table to the bios maybe a month after 45nm on the market.

[dnottis]

Interesting about P5K bios. Is that a recent change or has it been that way since 45nm launch in January?

The lowest PLL setting in the DFI bios is 1.55v. The VTT minimum I think is 1.10v or 1.12v. They added new 45nm GTL table to the bios maybe a month after 45nm on the market.

It has been that way since the bios started coming that supported 45nm. I believe 0906 was the first. 1004 and 1006 also have the same information within the bios and auto detects 45 or 65nm and adjusts accordingly. Maybe the DFI board and others are the reason why some people ARE having problems however. These boards that don't lower the PLL and VTT are just over juicing and frying 45nm chips.

[alpha0ne]

^ sounds like a good reason to put off buying a new mobo for a little while

[dnottis]

^ sounds like a good reason to put off buying a new mobo for a little while

I have strayed from Asus a few times, over the years they have put out a few bad boards too. But the p965, p35 have been flawless for me, with a simple pencil mod there is no droop. I always find myself coming back to Asus when I'm ready to get serious about my overclocking. I had a DS3R that was decent but once I dropped a quad in the vdroop made the board nearly unusuable. I dropped a e2160 in there for now.

[Fguarezi]

Q745A551

Hi.

Prime 4005 1.280v



There degradation using this?

[TerTop]

Im not sure if any of you have studied the erratas the mentioned in the updated white papers for E8000 series (released Feb 2008). It seems like the E8000 family of cpu got TONS of erratas. Maybe some of you semicon guys could shed some lights if any of this bugs are related to this "degration" issue or not.

http://www.intel.com/design/processo...pdt/318733.htm

[CrazyNutz]

Im not sure if any of you have studied the erratas the mentioned in the updated white papers for E8000 series (released Feb 2008). It seems like the E8000 family of cpu got TONS of erratas. Maybe some of you semicon guys could shed some lights if any of this bugs are related to this "degration" issue or not.

http://www.intel.com/design/processo...pdt/318733.htm

Yeah, I just read it. I found more than a couple that would/could cause programs like prime95 to fail, at stock speeds even! I have found a few
people claiming prime95 failing at stock speeds, and this is before ever
increasing voltage and/or overclocking the cpu (virgin cpu's).

Thanks for pointing out the errata document.

[Grnfinger]

Been running my wofldale at 4.3 or higher since the day I dropped it in my board.
I persoanally think this "degrading" is BS, altho I'm sure a 100 ppl will disagree







priming 4500 now so no shot of that

[Leeghoofd]

I persoanally think this "degrading" is BS, altho I'm sure a 100 ppl will disagree

Well I have 2 QX9650 CPU's that did about the same benching and co, and one realy has degraded for sure as the other still is working like it ran from day one... both used in the same setup , same WC setup , tralalala... some chips are good some are bad or get bad ...sadly to say...

But to say degradation is BS Grn that's a BS claim Thing is degradation can misunderstood with Mobo , ram failure too... but if I test on the same platform and I take notes of my CPU's required settings ( speed,voltage wise) and now one needs more voltage than before something must have gone bad....

[Grnfinger]

First I said in my opinion.....so to say my statement is BS is not valid.
If the cpu has "degraded" in any way why not RMA it?
If it was any other piece of hardware it would have been sent back long ago?
I dont doubt that there are perhaps "flawed" chips out there, but seroiusly how can a chip degrade? Please explain...Your own words are "some chips are good some are bad or get bad ...sadly to say... "
So lets say you had ram that worked great today, sketchy tomorrow and the third day, would you keep it and chock it up to degrading? Hell NO that ram would be RMA'd so why not the chip? ppl see what they want and for evey person that will agree with me there will be 100's that wont, Am I right and there wrong... NO, it means we have a difference of opinion and thats what makes this world go round.. diversity.
But if you spent all that cash on 2 QX chips and 1 is leaps and bounds better than the other send it back instead of slapping a "degrading" sticker on it

[Leeghoofd]

Well let me clarify more :

I firstly responded to your post because you used the magical BS ... seems you are pretty convinced that in your humble opinion some peeps claims are not correct... can be true, you cannot go and verify all the posts, some users just pump voltages and push stuff... things can break ofcourse... I'm happy for you your CPU holds up nicely... you got a good one cherish it....

But what I've seen and noticed is a first timer for me as it happened in the matter of a few weeks... and I've benched and abused CPU's before...never saw a CPU that needed so much more volts even to run at stock speeds...

Secondly Intel is advised about my issue, first I need to try to swap it via my shop , if no go Intel will handle the RMA, but Intel let it shine through in the reply that 1.19volts for 3ghz stock is not that bad...while the vid says 1.125...I also told them about 2 cores not reading out at all they stay put at 37°C... what do you think about that ? real bummer not ?...
So I'll reckon I'll bench it some more till it's really bad and then RMA it if it degrades more ofcourse...
Thirdly I need my Rig at home and the other one at my parents till at least June this year... so my shop can try with the retailer to get something arranged... if they give me a Quad on loan I will give them the QX straight away... and RMA in belgium is months of waiting mate, it's not just go in and swap sadly... and I already waited 3 months for my Qx's to be availble...

Some peeps seem to be able to run them over 1.4 volts easily, some suddenly get lockups, BSOD, reboots, I know at least 3 peeps with 45nm Qx quads that have similar issues to mine but ofcourse it might be something else than just the CPU in their case... I can test on my 2 rigs , swap hardware like that...
Sadly that troublesome CPU does the same thing on my 2 rigs... it just needs more volts than before, even at stock with nothing OC'ed... degradation : hell yeah !!! question remains why this one and not the other ?...

[dnottis]

My chip isnt as good as some of the others out there, temp sensors are stuck lol. Needs 1.39v for 4Ghz.

PLL and VTT are stock.

[MosIncredible]

^ I also have a Q748A219 and it takes me the same voltage for 4GHz stable. I run 3.9GHz 24/7 because I'm uncomfortable with that voltage.

[dnottis]

^ I also have a Q748A219 and it takes me the same voltage for 4GHz stable. I run 3.9GHz 24/7 because I'm uncomfortable with that voltage.

Mine is water cooled I can run 4050 @ 1.39v, 4.1 @ 1.41v, 4.2 @ 1.45v. Core temps have never broken 56 according to real temp, however my idle temps are always 50 so I don't really trust the temps, I'm betting they are lower. I dont mind 1.4 or 1.45v anything more makes me nervous with 45nm - also I can't say I've ever had a CPU more than 6 months. I'm happy with it, booted over 500 so no low FSB ceilings. Bought it used for $210, lapped it and been playing with it ever since.

Like most wish I had a magical chip that ran 4.2 Ghz @ stock volts though lol.