**Official e8400/e8500 Retail OC Thread

[TEDY]

where do you get this priceless chips? :S

[hersounds]

This processor was bought in Tokyo last February by a friend of mine.

from tokio to buenos aires

something like this

http://page8.auctions.yahoo.co.jp/jp/auction/h54341094

[mad_skills]

Guys, a piece of advice if you please.
I thought of swapping my E6750 for an E8400 so if you could give me some pointers on what FPO, pack date, revision i should look out for etc.
I guess C0 revision is the best one out right now, so what's the marking for it.
Tnx.

[brinox]

i would just take a look at this thread all the way through, noting batch numbers and stuff. usual intel rule(s) is the earlier the pack date on the same stepping is best.

from what ive read, the Q746A and Q750A are a pretty good choice, E8400 and E3110 alike.

I myself have an E3110 Q746A534 waiting at home for my replacement vanilla 680i SLI on friday... shooting for 450x9 until X48

[The Viper]

i would just take a look at this thread all the way through, noting batch numbers and stuff. usual intel rule(s) is the earlier the pack date on the same stepping is best.

from what ive read, the Q746A and Q750A are a pretty good choice, E8400 and E3110 alike.

I myself have an E3110 Q746A534 waiting at home for my replacement vanilla 680i SLI on friday... shooting for 450x9 until X48

dont forget the Q745A, which in my opinion has posted some of the best numbers

[MrHydes]

the diode temp and the idle/full temps work fine in E8400 ES or E8500?

[mk-ultra]

dont forget the Q745A, which in my opinion has posted some of the best numbers

my Q745A have failed to impress me, or i'm doing something wrong over here.

[mrcape]

Q745A576 has probably the best results I've seen anywhere in terms of hitting high stable clocks with very little voltage.

4ghz - 1.675vcore - SUperPi 1m
4ghz - 1.20vcore - Prime95
4.25ghz - 1.28cvore Prims, Orthos, OCCT stable
4.6ghz - 1.41vcore SuperPi 1m sub 10sec.

I dunno, has anyone seen better results?

[The Viper]

my Q745A have failed to impress me, or i'm doing something wrong over here.

Define "failed to impress" and thats why I said "some" but really thats pretty much the 1st Ive heawrd someone say about week 45

[WiTcHkInG]

Had a Q746a and a Q750a and both failed to hit 4Ghz.

Just got a Q748A402, will see if it goes well later. 3rd time lucky

[Mav451]

I believe dark22 had a pretty nice chip (Q748A133).

mrcape - any idea what it takes for just 4.3Ghz? I think you'd have the only chip here <1.3v for that frequency. I thought dark22 also did, but I have to check back a few pages.
Problem is that most on this thread are simply testing for max and throw gobs of voltage at it...which I don't think mad_skills would be interested in.

Had a Q746a and a Q750a and both failed to hit 4Ghz.

Just got a Q748A402, will see if it goes well later. 3rd time lucky

You need to include the last 3 numbers as well.

[mrcape]

Mav451 - Unfortunately I jacked the chip a bit up and I can't get that data from the perspective of a fresh, undamaged Q745A576 chip. I'm sure that before it was damaged it could do 4.3ghz with one more notch, maybe 1.29v. I say that based on the way it scaled up to 4.7 only needing 1.41v max.

All I can present for results are what I recorded before or after it degraded. 4.25ghz with 1.28v was the highest point where I stress tested the fresh chip. Now that same clock takes about 1.32v to be prime stable, but it is Priority 9 10k fft stable for hours. The reason for the drop is because I drank beer and beat on it one night trying to get the most out of some 3dmark. pfff

The gobs of voltage thing makes me cringe every time I see it! Some of those probably would have been stellar if treated properly. But to each his own you gotta do it to get the hwboints.

I'm already on the hunt for another Q745 still in the box unopened.

[WiTcHkInG]

You need to include the last 3 numbers as well.

Sold on the Q746A but think the last 3 were 366.
The Q750A is Q750A032

[Spyrus]

When comparing you should remember also the cooling used.
Water may give you 100-150MHz or more for example.
And of course orthos stable.
On air some chips like mine can do 3.167@1.3V for 32m but that is not the case.

[Falkentyne]

Mav451 - Unfortunately I jacked the chip a bit up and I can't get that data from the perspective of a fresh, undamaged Q745A576 chip. I'm sure that before it was damaged it could do 4.3ghz with one more notch, maybe 1.29v. I say that based on the way it scaled up to 4.7 only needing 1.41v max.

All I can present for results are what I recorded before or after it degraded. 4.25ghz with 1.28v was the highest point where I stress tested the fresh chip. Now that same clock takes about 1.32v to be prime stable, but it is Priority 9 10k fft stable for hours. The reason for the drop is because I drank beer and beat on it one night trying to get the most out of some 3dmark. pfff

The gobs of voltage thing makes me cringe every time I see it! Some of those probably would have been stellar if treated properly. But to each his own you gotta do it to get the hwboints.

I'm already on the hunt for another Q745 still in the box unopened.

Eh? What vcore and mhz caused it to degrade? I thought anything below 1.4v load was supposed to be safe?

[~aoe~]

Had a Q746a and a Q750a and both failed to hit 4Ghz.

Just got a Q748A402, will see if it goes well later. 3rd time lucky

That's the exact same batch as I've just received. Unfortunately, I won't be in a position to test it for another couple of weeks. Let me know how it goes. Good Luck ..

[mrcape]

Eh? What vcore and mhz caused it to degrade? I thought anything below 1.4v load was supposed to be safe?

I had it up to 1.41 - 1.42vcore max and mine got bruised. There's no clear answer to what can cause it. I'm guessing a combo of High FSB + highish vcore. There are many variables - board, other voltages, fsb etc..

[mk-ultra]

Define "failed to impress" and thats why I said "some" but really thats pretty much the 1st Ive heawrd someone say about week 45

the good> it does 3600 @ 1.19/1.20 (windows)

the bad> i'm having a hard getting it stable around 4ghz, right now i'm running 480x8.5 - 1.4125bios/1.4windows and i think i'm ok, it's prime stable(haven't tested for long half an hour only)but when i play cod4 it keeps crashing the game below that.

and i can forget about 500+fsb, maybe it's my mobo?

it's a Q745A553 pack date>12/27/07

[flesheatinvirus]

My 745A576 is still doing good on water after 7 weeks of somewhat harsh play. 4300mhz needs 1.375v to be solid and 1.465v for 4500, 4800 is still very bench stable with a max of just over 4900.

[Hughyhunter]

My 745A576 is still doing good on water after 7 weeks of somewhat harsh play. 4300mhz needs 1.375v to be solid and 1.465v for 4500, 4800 is still very bench stable with a max of just over 4900.

What are the rest of your voltages (ie NB, SB, FSB)?

Ram linked or unlinked at what settings?

What mobo?

[flesheatinvirus]

For example running at 4300mhz on my P5K Prem board needs 1.375v CPU,1.60PLL,1.30FSB term,1.40NB,1.05SB,Clock Overcharge .90,Transaction Booster disabled,CPU&NB GTL is .59 and .61. Ram is Ballistix at 1148 4-4-4-12-3-35-6-3-3 at 2.35v on a 333 strap with static read enabled.

[Xvys]

I have the P5K Deluxe, flesheatinvirus, and I notice you have the Transaction Booster disabled. With the newest bios 809, with TB disabled the Performance Level has dropped to 9, while it was at Level 6 with the older bios when disabled. This is a big performance hit. Level 6 has about 20&#37; lower latencies than Level 9 on my system. I would assume the similar Premium board has the updated bios available now also.

I am running at 4005mhz on my P5K Deluxe board with 1.365v cpu in bios,1.50PLL,1.10FSB term,1.25NB,1.05SB,Clock Overcharge disabled,Transaction Booster enabled 2 (Performance Level 6),CPU&NB GTL is Auto and .61. Ram is Ballistix at 1112 MHz 5-5-5-12-2-25-5-2-2 at 2.10v on a 333 strap with static read disabled.

[Falkentyne]

I had it up to 1.41 - 1.42vcore max and mine got bruised. There's no clear answer to what can cause it. I'm guessing a combo of High FSB + highish vcore. There are many variables - board, other voltages, fsb etc..

Thanks.
Did you have it at this vcore for several hours or all night, or only briefly (for a few minutes?).

[ThatGuy16]

If coretemp is even remotely correct. is 60-63c well within recommended temps? This is only at 1.28v for 4.0ghz. The "CPU" temp however reads 45ish on load. i have never seen the core and cpu temp that far apart, even though they are on difference sensors.

[Spyrus]

If coretemp is even remotely correct. is 60-63c well within recommended temps? This is only at 1.28v for 4.0ghz. The "CPU" temp however reads 45ish on load. i have never seen the core and cpu temp that far apart, even though they are on difference sensors.

I have similar results with air and 1.3V@4050MHz and cpu is at 46-47 and is orthos stable (on Asus Maximus Formula) so i would not worry.