**Official e8400/e8500 Retail OC Thread

[fng77]

Should be really easy to hit 4G.
Bios settings and ram timings can make a huge difference in stability.
Go easy on the voltages.

And congratulations!

Thanks!

Cant wait. I wonder how much more sensetive the 45 nm cpu is on vcore, my 2160 needs around 1,525 for 3,6 ghz, but then it gets very hot, up to 80 c (coretemp) under orthos with my xigmatek.

[Obitus]

CPU: E8400
Batch #: Q747A374
Ship/Packaging Date: 02/05/08
Voltage(s) (Stock/OC'ed): 1.22 / 1.425
Frequency(ies) (OC'ed): 4.2Ghz 9 x 467
Motherboard: GA-EX38-DS4
BIOS: F3
-----
*Extra* Memory(ies) (Timing(s)) Stock/OC: 5-5-5-16 / 5-7-7-22
*Extra* Memory(ies) (Voltage(s)): 2.034
*Extra* Third Party Cooling Solution: Xigmatek HDT-S1283
*Extra* Paste(s)/Compound(s)/Other(s): Tuniq TX-2

I can boot and run SuperPI and other apps in both 4.2 and 4.05 but I can't pass Orthos for more than a few seconds. I can run at 3.6Ghz & 1.237v all day long. I know there are other settings to play around with but just don't know enough right now to feel comfortable doing it.

[fng77]

CPU: E8400
Batch #: Q747A374
Ship/Packaging Date: 02/05/08
Voltage(s) (Stock/OC'ed): 1.22 / 1.425
Frequency(ies) (OC'ed): 4.2Ghz 9 x 467
Motherboard: GA-EX38-DS4
BIOS: F3
-----
*Extra* Memory(ies) (Timing(s)) Stock/OC: 5-5-5-16 / 5-7-7-22
*Extra* Memory(ies) (Voltage(s)): 2.034
*Extra* Third Party Cooling Solution: Xigmatek HDT-S1283
*Extra* Paste(s)/Compound(s)/Other(s): Tuniq TX-2

I can boot and run SuperPI and other apps in both 4.2 and 4.05 but I can't pass Orthos for more than a few seconds. I can run at 3.6Ghz & 1.237v all day long. I know there are other settings to play around with but just don't know enough right now to feel comfortable doing it.

Sounds great! 1.237 vcore is what I need for 2,4 ghz (stock 1,8 ghz). how much vcore do you need for 4.2 ghz and 4.05?

[fng77]

And what are your temps at that speed?

[Obitus]

Sounds great! 1.237 vcore is what I need for 2,4 ghz (stock 1,8 ghz). how much vcore do you need for 4.2 ghz and 4.05?

Heh, it says right in my results. look again 1.425v for 4.2. What it didn't show is 1.375v for 4.05. Load temps for 4.2 are about 62c and for 4.05 is 59c.

[CrazyNutz]

Just picked up a e8400 at frys for $199 batch q747a258 any luck with this batch?? Also looking for tips on a solid board to get 4.0 on? Thanks guys.

I got a 747A377 Pack Date:02/05/08 from Fry's and It is not that great
It needed way to much voltage to get over 4ghz (>1.4v) and was not that
stable at that. I also got a 748A144 which does much better but gets a little
warmer.

[fng77]

Heh, it says right in my results. look again 1.425v for 4.2. What it didn't show is 1.375v for 4.05. Load temps for 4.2 are about 62c and for 4.05 is 59c.

Sorry. I can se that now.

But anyways, Great results I think. 4.2 at 1.425 isnt bad at all. I just hope I get a good cpu too. I could definetly live with 1.375 vcore for 4.05 Ghz.

I´m at 1.35 right now, thats what I need for 334*9 fsb (3006 Mhz)

[fng77]

Nice that you have a xigmatek too, then I really have somone to compare my results with. I`m gonna try to fix the money until next week, the cpu is lying waiting for me at the shop.

Found a demo ex. for sale at a hardware store here in Sweden. I got it for just 200 USD thats pretty cheap for being here.

[pervi]

don't worry about differences unless it's causing stability issues

ok, thks

do you think is the reference temperature? witch core ?

[Grnfinger]

4.3GHz @1.38 15 hours prime stable.
4.4GHz @1.40 8 hours prime stable.
4.5GHz @1.45 8 hours prime stable.

All tests were done running BLEND, I run 4.3GHz for 24/7 and have never seen the "degrading" issue. I have ran this chip at atleast 4.3 right out of the box. Performance, benchmarks and stability have never degraded. With new beta bios 1004 I have actually gained a few 100 points on 3dmark

[alcatraroz]

Sorry. I can se that now.

But anyways, Great results I think. 4.2 at 1.425 isnt bad at all. I just hope I get a good cpu too. I could definetly live with 1.375 vcore for 4.05 Ghz.

I´m at 1.35 right now, thats what I need for 334*9 fsb (3006 Mhz)

that s not so great result its an average result for those cpu..i stay daily (orthos >2hr stable ) at 4.05 with just 1.29v

[ThatGuy16]

What do you guys think is a safe voltage on air, only for bench's such as 3dmark06?

For 4.3ghz in 3dmark, it needs 1.4v in bios and is 1.36v in windows. Is it safe to go higher? i don't want to hurt my chip

i have always told myself max 1.4v.. but i want to go higher

[fng77]

that s not so great result its an average result for those cpu..i stay daily (orthos >2hr stable ) at 4.05 with just 1.29v

Great!

[CrazyNutz]

What do you guys think is a safe voltage on air, only for bench's such as 3dmark06?

For 4.3ghz in 3dmark, it needs 1.4v in bios and is 1.36v in windows. Is it safe to go higher? i don't want to hurt my chip

i have always told myself max 1.4v.. but i want to go higher

Since you dont want hurt your chip I would stay at the 1.36 (in windows).

[tranceaddict]

When Intel specifies the max operating voltage of 1.3625 and absolute max of 1.45V; are they refering to the voltage inputed in BIOS or the actual measured voltage?

Thanks

[leffe]

When Intel specifies the max operating voltage of 1.3625 and absolute max of 1.45V; are they refering to the voltage inputed in BIOS or the actual measured voltage?

Thanks

Actual measured voltage ofcourse.

[massmint]

I built a new rig recently and I'm content with what I can do with it. My last rig was a MSI K8N-Neo4 Sli Platinum with X2-3200 OCed 3.2 Water (Zalman block) + Termaltake Bigwater Kit(Without cpu block, It cracked). This is the new rig OCed on Air.

[.OCX]

Quite simply, we believe any voltage over 1.45V is asking for trouble with 45nm processors and our conversations with Intel to date have all but confirmed our suspicions.

- Anandtech Article

The article also talks about Degradation and Warranties and how this affects a product life cycle. The important thing is the cross relationship between high temperatures and high voltage over long periods of time.

On the other hand, perhaps precautions will be taken in order to accommodate the higher voltages like the use of water-cooling or phase-change cooling. In any case, the underlying principles are the same - overclocking is never without risk. And just like life, taking calculated risks can sometimes be the right choice.

- Anandtech Article

Great article, only thing it lacked was ANY mention of how FSB Termination Voltage effects overclock stability, degradation, etc...Seems no one wants to be very forthcoming with 45nm Fabs and FSB Term(VTT) voltages differences from 65nm Fabs.

What I'm taking away from this article:
By Overclocking, there is a risk I could damage my 45nm CPU, but by Watercooling it now becomes a calculated risk because the cooling properties of Water have reduced my risk of damaging my CPU with high voltage and temps compared to those who choose Aircooling methods and use exceedingly higher voltage values and temps to achieve the same results.

[ThatGuy16]

Since you dont want hurt your chip I would stay at the 1.36 (in windows).

Thanks

so you guys think ~1.36v is the highest only for benches? (on air)

[Mav451]

What I'm taking away from this article:
By Overclocking, there is a risk I could damage my 45nm CPU, but by Watercooling it now becomes a calculated risk because the cooling properties of Water have reduced my risk of damaging my CPU with high voltage and temps compared to those who choose Aircooling methods and use exceedingly higher voltage values and temps to achieve the same results.

What is so different with water cooling? I mean is it that drastically lower in temps? Or is it different in the way heat is moved away:
Is it b/c the heat is dissipated more quickly?
Smoother b/c it's a moving liquid instead of air?

And what does air cooling have to do with "exceedingly higher voltage values"? Why would someone with air be using more voltage than a water cooled CPU?

[.OCX]

What is so different with water cooling? I mean is it that drastically lower in temps? Or is it different in the way heat is moved away:
Is it b/c the heat is dissipated more quickly?
Smoother b/c it's a moving liquid instead of air?

And what does air cooling have to do with "exceedingly higher voltage values"? Why would someone with air be using more voltage than a water cooled CPU?

Theres a lot of sticky threads on this forum that explain most of your questions with regards to aircooling vs. water. But for the 45nm example:

Typically...
4.0ghz:
Aircool 1.35v-1.40v
Temps: 60c+ Loaded

4.0ghz:
H20 Cool: 1.26-1.30v
Temps: 40-50c Loaded

This is generalized data that I have noted on countless Orthos/Prime/OCCT stability screenshots that folks have graciously posted. The 45nm Database you put together, although valiant, unfortunately had several flaws which invalidated most of the recordings. Specifically, the stability testing/methodology and the validity of each user's admission. Without that control measure in place, the screenshots of stability testing is really all there is to go on.

Sure, there will be those who say "But I can achieve 1.324v on Air to hit 4.0ghz" or something along those lines. There will always be standard deviations when collecting data, and especially when generalizing data to make inferences based upon those conclusions. But the fact remains the same, using more energy to achieve the same results as someone who uses less, ceteris paribus, is an inefficient use of resources, and couldn't possibly be adopted as a "Best Practices" policy or guideline for scientific use. Not to mention the "Calculated Risk" factor involved with 45nm technologies.

[joe.denice]

i have an e8500 that does 4ghz (8x500) w/ 1.275 volts... How is that?? I will post some screenies later.

[Mav451]

Something doesn't sound right about that. Just b/c you use a different method of cooling, doesn't mean that the processor isn't going to overclock as well. I think the other factor you aren't considering is that they are dramatically less people using WCing. And yeah, my database has bad data b/c I didn't set a standard for the OCs. What I should have set as the baseline limit is 1hr OCCT - which eliminates those useless SuperPI OCs.

I think they are many users who've already posted in this thread that have 4Ghz with <1.3v - and all of us are on air. I still think the biggest deal with these chips is a good batch, not your method of cooling. Frankly, concluding that b/c you're using WCing guarantees a .09v drop in required vcore is just plain bogus. I understand you like WCing, but to make a claim like that is just ridiculous.

[HeinzS]

Theres a lot of sticky threads on this forum that explain most of your questions with regards to aircooling vs. water. But for the 45nm example:

Typically...
4.0ghz:
Aircool 1.35v-1.40v
Temps: 60c+ Loaded

4.0ghz:
H20 Cool: 1.26-1.30v
Temps: 40-50c Loaded

please have a look in this thread: http://forumdeluxx.de/forum/showthread.php?t=457105

[Tuvok-LuR-]

my E8200 Q748A230


cooling: Freezer 7 Pro (Air)
ram: 2x1GB Teamgroup Elite PC6400
mb: GA-P35-DS3L
psu: corsair 450vx