**Official e8400/e8500 Retail OC Thread

[zakelwe]

I think we need to stop worrying about temps until people start reporting not being able to get around throttling problems. Has anyone reported throttling problems yet?

Of more concern is the wall that seems to be hit at a certain frequency where an increase in volts will not get you anything more. Both myself and Zeus have hit this at 4600Mhz, has anyone else seen such a thing?

Normally you would expect more volts to give you a bit more but this seems to be a proper wall. This wall seems a lot more substantial than any effects caused by high temps.

Comments appreciated.

Regards

Andy

[Jeremy617]

as in you can't post past 4600mhz or just can't get into windows?

[Zucker2k]

Originally Posted by zakelwe

I think we need to stop worrying about temps until people start reporting not being able to get around throttling problems. Has anyone reported throttling problems yet?

Of more concern is the wall that seems to be hit at a certain frequency where an increase in volts will not get you anything more. Both myself and Zeus have hit this at 4600Mhz, has anyone else seen such a thing?

Normally you would expect more volts to give you a bit more but this seems to be a proper wall. This wall seems a lot more substantial than any effects caused by high temps.

Comments appreciated.

Regards

Andy

The buck has to stop somewhere.

What I find intriguing is that some of the chips seem to have a dynamic vid. Below is a picture of P95 running at 4.6Ghz on x38 and the vid. is still 1.1125 underload. So, my question is, is the this (problem?) related to mobos of a particular chipset? Also, the lower your vid, the higher your temps are going to be so it'll seem to me that those reporting dipping vids on load are probably the same people reporting high temps. Please keep an eye out.

Another thing, The latest everest .....270, has a pretty good stability testing systems which slowed down my system to a crawl, but most importantly, has a cpu throttling indicator; it should be good for anyone whose temps. typically reach the high 80s to try and use that, or merely have it open while orthos runs to determine if there's any throttling taking place. I think RMClock has the feature too. Cheers!

[loc.o]

Originally Posted by zakelwe

I think we need to stop worrying about temps until people start reporting not being able to get around throttling problems. Has anyone reported throttling problems yet?

Of more concern is the wall that seems to be hit at a certain frequency where an increase in volts will not get you anything more. Both myself and Zeus have hit this at 4600Mhz, has anyone else seen such a thing?

Normally you would expect more volts to give you a bit more but this seems to be a proper wall. This wall seems a lot more substantial than any effects caused by high temps.

Comments appreciated.

Regards

Andy

Simular results except for the wall at 4600. My E8500 does all 3Dmarks at 4510mhz with 1.48v however not over 4560mhz in 1M same volts. Never get an error because of cpu speed spi just shuts down. Goin up to 1.60v gains me about 150mhz for 1M though, max about 4710mhz at those volts.

[GLENBOY]

same here at 4600 mhz it just stops really did aquamark at 4660 but at 4500 you can do anything don't know whether its a fsb thing on x38 i have e8400 but e8500 was the same

[Merc14]

You have to say that it is unbelievable that any of us are complaining about hitting a wall at 4600MHz. Just a year and a half ago breaking 3000MHz on an Opty 165 was a huge achievement. Coretemp issues aside, this is one amazing CPU.

[GLENBOY]

good point mate we are just greedy

[ThatGuy16]

I ordered mine from mwave and a x38-ds4 from the egg. Hopefully it will all be here by wednesday

Now im worried about this temp and degrading thing .. maybe i'll just stop a 4ghz and make it prime stable.

Can't wait for my results though

[nama]

Originally Posted by avrionis

I had a e8400 had the same results.. Had great temps! But past 4.2...I would immediately get those errors on core2. I had tried many different voltage combinations but no luck.. Sent it back..

did you get a new one? if so any success?

[zakelwe]

Originally Posted by GLENBOY

same here at 4600 mhz it just stops really did aquamark at 4660 but at 4500 you can do anything don't know whether its a fsb thing on x38 i have e8400 but e8500 was the same

I don't think it is an X38 thing, I have P35.

Like you I can do 3d all night at 4500 but get it up to 4600 and it will not do SuperPI 1M. I know my memory does 600+ and the motherboard 510 but at 484x9.5 onwards it's DEATH VALLEY !

As for the person saying "well 4.5GHz is really good anyway" I totally agree. For the amount I paid for this cpu it knocks my socks off.

However it is very interesting how these new process Wolfdales behave and if we can work out

a) the temp anomalies
b) possible walls at certain Mhz

then we can perhaps bend these cpus more to our whim than we are at present. It's a big learning experience.

I did notice that if I knocked off all the BIOS voltages for FSB and northbridge etc by 0.1v-0.2v that I could still do 484FSB and my cpu temp went down by approx 5C under load. I still hit the wall though.

Regards

Andy

[HaTaX]

Okay update on where I'm at. I did the vdroop mod as explained by JohnnyBravo using a 1Kohm variable resistor. My VCore now stays within .002-.003 volts of where I set it at idle or load. There is a little flutter there, but it's nothing bad and it is certainly MUCH better then it was before.

This helped get down a little on VCore required to reach certain clocks, but it didn't fix the fact that my E8400 still takes more voltage to reach some of the clocks people here are seeing.

I called over to Microcenter where I got the cpu and mentioned the temp problems with the CPU and the person said they'd be happy to exchange it for me, but right now they didn't have any of them left in stock.

Soo, I'll call over there in a day or so and ask if they've gotten any in, and if so, I think I'll do an exchange!

Is there a table being kept track of anywhere that reports batch numbers and what people have gotten with them?

[dinos22]

Originally Posted by loc.o

Simular results except for the wall at 4600. My E8500 does all 3Dmarks at 4510mhz with 1.48v however not over 4560mhz in 1M same volts. Never get an error because of cpu speed spi just shuts down. Goin up to 1.60v gains me about 150mhz for 1M though, max about 4710mhz at those volts.

yeah my wall is 4700 lol

it's very interesting

obviously with SS it flies way past that

[TSFroggy]

Same stuff in sig, with the following specifics:
FPO/Batch: Q746A377



Currently priming for around 40 minutes. Voltage set in BIOS is ~1.36v, after droop CPU-Z says 1.328v. It wouldn't prime for anything around 4050, but this seems to have more luck.

I'll have more voltages for you when I get around to restarting.

Mem is at 5-5-5-18 471MHz 2v. Nothing special... yet.

(Almost didn't find a print screen, till I realized that F14 is print screen on Apple keyboards - I love this thing! :P)

[MAXLD]

Which cooler are you using?

[chrisf6969]

Preliminary results after only an hour of playing:


http://valid.x86-secret.com/show_oc.php?id=306969

I'll tweak/play some more this weekend and hopefully hit 4Ghz.

[Bail_w]

@Hazardous - each chip has its own potential, so dont expect the same results everytime even though it is the same batch.

[eva2000]

Originally Posted by zakelwe

I think we need to stop worrying about temps until people start reporting not being able to get around throttling problems. Has anyone reported throttling problems yet?

Of more concern is the wall that seems to be hit at a certain frequency where an increase in volts will not get you anything more. Both myself and Zeus have hit this at 4600Mhz, has anyone else seen such a thing?

Normally you would expect more volts to give you a bit more but this seems to be a proper wall. This wall seems a lot more substantial than any effects caused by high temps.

Comments appreciated.

Regards

Andy

similar my E8500 Q740A493T max out on water around 4653mhz at 1.6v for 3dmark http://i4memory.com/showthread.php?p=83929#post83929 and around 4714mhz for super pi 32m on water.



Could be heat related as on 1/4HP SS it does scale higher

32M passed at

• @4851Mhz at 1.584v idle (0.08v droop under load)
• @5007Mhz at 1.680v idle (0.08v droop under load)
• @5033Mhz at 1.704v idle (0.08v droop under load)
• @5120Mhz at 1.808v idle (0.08v droop under load)
• @5155Mhz at 1.824v idle (0.08v droop under load)

[scwam]

TSFroggy, we have the same batches (Q746A377 1/4/08). Did you get yours from TankGuys?

I'll update this pic with longer stability as it goes by
VID is low, but yet I feel the temps are good. Just got this chip today. Thermistor I put under the socket is reporting temps concurrent with previous systems I've used it on about 11c lower than what Coretemp reports at load. I'm not concerned about the temp issue since I've got my own device. This chip runs cooler than my 6850.

Water 1.414 DMM verified load measurement:

[zakelwe]

Woohoo! After 3 hours of struggle I managed sub 10 second using a stock Intel heatsink fan

http://www.zen141854.zen.co.uk/9985.jpg

http://www.zen141854.zen.co.uk/intel.jpg

Slight cheat as it is off an E4500 )

I am the King of Pi ! Should change my name to K|ngP| .

That looks like a good name for a bencher

Regards
K|ngP|

[syne_24]

Yea there is definitely a wall after 4.6 for 3DMark. I can run @ 4.725 but it hangs on the CPU test. As for Pi, I really can't confirmed since running on vista x64 already has issues with it. So far it looks like 4.7 is about max on water @ 1.6v (bios) and probably 4.3~4.4 for 24/7 with some fine tuning.

[TEDY]

zakelwe intel stock cooler with blue fan, is it with CU base?

[yenclas]

Originally Posted by Zucker2k

The buck has to stop somewhere.

What I find intriguing is that some of the chips seem to have a dynamic vid. Below is a picture of P95 running at 4.6Ghz on x38 and the vid. is still 1.1125 underload. So, my question is, is the this (problem?) related to mobos of a particular chipset? Also, the lower your vid, the higher your temps are going to be so it'll seem to me that those reporting dipping vids on load are probably the same people reporting high temps. Please keep an eye out.

Another thing, The latest everest .....270, has a pretty good stability testing systems which slowed down my system to a crawl, but most importantly, has a cpu throttling indicator; it should be good for anyone whose temps. typically reach the high 80s to try and use that, or merely have it open while orthos runs to determine if there's any throttling taking place. I think RMClock has the feature too. Cheers!

Where is this beta version for download ?

Thanks !

[Zeus]

Originally Posted by zakelwe

Woohoo! After 3 hours of struggle I managed sub 10 second using a stock Intel heatsink fan

http://www.zen141854.zen.co.uk/9985.jpg

http://www.zen141854.zen.co.uk/intel.jpg

Slight cheat as it is off an E4500 )

I am the King of Pi ! Should change my name to K|ngP| .

That looks like a good name for a bencher

Regards
K|ngP|

Well done K|ngp|!!!

[dinos22]

Originally Posted by zakelwe

Woohoo! After 3 hours of struggle I managed sub 10 second using a stock Intel heatsink fan

http://www.zen141854.zen.co.uk/9985.jpg

http://www.zen141854.zen.co.uk/intel.jpg

Slight cheat as it is off an E4500 )

I am the King of Pi ! Should change my name to K|ngP| .

That looks like a good name for a bencher

Regards
K|ngP|

now try doing that at 4.5GHz even

[TSFroggy]

Originally Posted by MAXLD

Which cooler are you using?

Tuniq Tower 120. Boy, this proc sure runs hot. :P

Originally Posted by scwam

TSFroggy, we have the same batches (Q746A377 1/4/08). Did you get yours from TankGuys?

I'll update this pic with longer stability as it goes by
VID is low, but yet I feel the temps are good. Just got this chip today. Thermistor I put under the socket is reporting temps concurrent with previous systems I've used it on about 11c lower than what Coretemp reports at load. I'm not concerned about the temp issue since I've got my own device. This chip runs cooler than my 6850.[/IMG]

Good results there scwam, and yes I got it from TankGuys. Looks like I only have an average chip to my name - it keeps asking me for more and more volts to the vCore, and I can't do that till I know what's safe for 24/7. Especially considering the droop on this board - I really need a vDroop mod. :P