View Full Version : E6750 3.9Ghz on Air, Prime Stable!!!
milan03
12-15-2007, 05:53 PM
I've been trying to squeeze the maximum stable clock out of this chip for days, and therefore the best bang for the buck, and the absolute top I could do with this setup is 3.9Ghz prime stable. I can boot @ 500x8, but it hangs just after few seconds in orthos. THis is almost 50% OC for E6750 on air, and in my book this is incredible.
Temps are also fine around 60C @ 1.57V in bios, vdimm is 2.2V, NB is up 2 notches (VTT Max), SB up 1 notch, GTLREF is max.
Will be happy if this chip lasts me for 2-3 years, and then I'll probably build a new system. I mainly use this for Hi Def HTPC so it should be sufficient for a while.
Thanks for everyone that left really helpful comments on this forum that I've been reading and learning.
Cheers :)
RPGWiZaRD
12-15-2007, 06:27 PM
Nice result, a lot better than mine that does 3.75GHz 1.52v. Best E6750 I've seen though does 4GHz 1.52~1.55v or so stable on air if I remember right. But it still a nice result as average seems to be 3.7~3.8GHz around that voltage. This CPU sure has a great value/performance ratio and therefore I decided to pick up one until Penryn arrives, can't resist the cheapish Wolfdale E8400. :)
I don't know how your temps can be that great though, I have like 65C in Orthos and not using a comp case.
milan03
12-15-2007, 06:42 PM
Room temp was at about 19C, and also I'm using u-120 Extreme. That thing is a beast! :D
RPGWiZaRD
12-15-2007, 06:58 PM
Well in average TRUE 120 provides about 3~4C better load temps in different reviews than Tuniq Tower 120 with same fan so it's not that big difference. Room temp 19C lol? It's winter here and around 0C or slightly on the subzero side usually and I have a lot warmer in this room despite no elements are on prolly like 23C... The blody comp that stays on 24/7 just warms up the room too much. :D
milan03
12-15-2007, 07:05 PM
I like to have the room at around 20C cause the damn heater in here is either on or off, and when its on, its too hot even with no clothes :) So i have the heaters on in other rooms, and keep it off in the computer room :)
The good thing is your chip could do that high FSB. ;) I never had a cooling problem, 3.6GHz was 1.35V fully stable, mine could not although do above 3.784GHz fully stable at 1.435V and max unstable 3804MHz at the same volts, but when you have locked multi's you have to have high FSB so get the drawbacks. So there is no way the chip could do anymore on FSB, so no higher MHz despite temps/volts.
What's your batch and VID?
Your temps are much higher than mine but mine was ambient 0C, 4C air blowing on it at over 2x 250CFM and 1x 500CFM 20inch... :D
And it was half outside the window to the ground at times where it was -4/-8C. Even sealed the chip onto a compressor for 30 minutes and then threw it in just just to see if it made any difference or if going colder would help. Nope, that was chip FSB limit.
milan03
12-15-2007, 07:24 PM
The good thing is your chip could do that high FSB. ;) I never had a cooling problem, 3.6GHz was 1.35V fully stable, mine could not although do above 3.784GHz fully stable at 1.435V and max unstable 3804MHz at the same volts, but when you have locked multi's you have to have high FSB so get the drawbacks. So there is no way the chip could do anymore on FSB, so no higher MHz despite temps/volts.
What's your batch and VID?
Your temps are much higher than mine but mine was ambient 0C, 4C air blowing on it at over 2x 250CFM and 1x 500CFM 20inch... :D
And it was half outside the window to the ground at times where it was -4/-8C. Even sealed the chip onto a compressor for 30 minutes and then threw it in just just to see if it made any difference or if going colder would help. Nope, that was chip FSB limit.
Batch# 3741B003 not sure about VID (where to locate that)
I first bought e2160 M0, but it had terrible fsb wall 320Mhz, so I replaced it with this one, and OMG what a difference... Too bad the multi is only 8x
CrayS-per4m
12-15-2007, 08:17 PM
thats a nice setup, i was running my E6750 at 3.85GHz on orthos for 3 hours yesterday and it was fine and it was at 1.512 volts on a 680i , i wanna hit 4.0 though lol im also using the Ultra 120Extreme but i have better load temps than you, underload i dont get above 53C but my case is a Cooler Master that's all metal mesh ... what fan are you using on the 120 extreme, i have a 133CFM Scythe Ultra Kaze, it works amazing!
milan03
12-15-2007, 08:43 PM
thats a nice setup, i was running my E6750 at 3.85GHz on orthos for 3 hours yesterday and it was fine and it was at 1.512 volts on a 680i , i wanna hit 4.0 though lol im also using the Ultra 120Extreme but i have better load temps than you, underload i dont get above 53C but my case is a Cooler Master that's all metal mesh ... what fan are you using on the 120 extreme, i have a 133CFM Scythe Ultra Kaze, it works amazing!
Scythe S-Flex Fan - SFF21E. Trying to keep it as silent as possible since my rig is ultimately an HTPC. but I'm happy with the temps at 1.57v in bios.
disruptfam
12-15-2007, 08:50 PM
i have a e6750 sitting on my desk not doing anything...
do you guys think i should swap it out with q6600 (b3)???
i can get my q6600 to 3.4ghz....on water...
do you rekon i could get 4ghz with my e6750 on water??
cheers all
milan03
12-15-2007, 08:53 PM
i have a e6750 sitting on my desk not doing anything...
do you guys think i should swap it out with q6600 (b3)???
i can get my q6600 to 3.4ghz....on water...
do you rekon i could get 4ghz with my e6750 on water??
cheers all
Why don't you try and see. I dont see why not...
disruptfam
12-15-2007, 09:52 PM
ohh sorry i forgot 2 ask would having a e6750 @ 4ghz beat out @ q6600 @ 3.4ghz???
i'm thinking the dual would be better in gaming but the quad still would be better in encoding/decoding etc
CrayS-per4m
12-15-2007, 09:52 PM
i have a e6750 sitting on my desk not doing anything...
do you guys think i should swap it out with q6600 (b3)???
i can get my q6600 to 3.4ghz....on water...
do you rekon i could get 4ghz with my e6750 on water??
cheers all
you definetly could if we are doing 3.8-3.9 stable on air
milan03
12-15-2007, 10:02 PM
Ive had GTLREF setting maxed out. Should I bring that setting back to default?
milan03
12-16-2007, 04:37 AM
The good thing is your chip could do that high FSB. ;) I never had a cooling problem, 3.6GHz was 1.35V fully stable, mine could not although do above 3.784GHz fully stable at 1.435V and max unstable 3804MHz at the same volts, but when you have locked multi's you have to have high FSB so get the drawbacks. So there is no way the chip could do anymore on FSB, so no higher MHz despite temps/volts.
What's your batch and VID?
Your temps are much higher than mine but mine was ambient 0C, 4C air blowing on it at over 2x 250CFM and 1x 500CFM 20inch... :D
your volts are much lower than mine also. I'm @ 1.57V in BIOS.
@ 3.6 I'm pretty much at the default voltages everywhere, and the temps hardly hit 40C full load after hours of orthos.
i'm all about pushing this baby 24/7 :D
Yeah the chip I had didn't need volts until you reached its FSB limit. 3720 and above needed more volts but 480 was max possible FSB IIRC and that could not change no matter if you gave it stock volts or max volts (1.6V-1.9 for all). Top 24/7 stable was 3720-3750MHz. But the boards I tried had large vdrops/vdroops too. Unfortunately my DMM crapped out and started reading false values and so I could not really test accurately any volt readings or mod them.
milan03
12-16-2007, 11:47 AM
Yeah the chip I had didn't need volts until you reached its FSB limit. 3720 and above needed more volts but 480 was max possible FSB IIRC and that could not change no matter if you gave it stock volts or max volts (1.6V-1.9 for all). Top 24/7 stable was 3720-3750MHz. But the boards I tried had large vdrops/vdroops too. Unfortunately my DMM crapped out and started reading false values and so I could not really test accurately any volt readings or mod them.
not sure what should be my highest temperature on full load... 70C or 75C? Cause in that case i could give it 1.65V+ and try 4ghz since my temps right now are not pushy at all...
I maybe the wrong person to query about that.. :D
I don't tolerate or advise running above 55C 24/7 for C2D/C2Q. My B3 C2Q used to start giving errors after 54C.
For testing max valids, I never go above 68C (benching basically).
But there's not one valid yet where I've been even 45C idle with Core 2 CPU's although load I've been to 60C max. I've always been an advocate of running as cool as possible with minimum cost and time wasting.
LennyRhys
12-16-2007, 04:12 PM
Very nice work. :cool: I can bench SPi and do various other things at 3.9GHz/1.425v without any bother (and with intel HSF lol), but stressing my CPU will commence after I have lapped the IHS and intsalled my new cooler from santa...and at 4GHz. :D
milan03
12-16-2007, 04:19 PM
Very nice work. :cool: I can bench SPi and do various other things at 3.9GHz/1.425v without any bother (and with intel HSF lol), but stressing my CPU will commence after I have lapped the IHS and intsalled my new cooler from santa...and at 4GHz. :D
how did you manage to get the 4ghz stable at only 1.425? is that value in Windows or in BIOS? I assume at that voltage you cant really stress test.
See I never can figure out if I should even run stress tests since I dont really game or fold, all i do is watch HI Def movies, and I don't think that will stress my cpu this much. I can totally superPi at over 4Ghz, but it does require me to up the voltage to over 1.61V.
LennyRhys
12-16-2007, 04:38 PM
how did you manage to get the 4ghz stable at only 1.425? is that value in Windows or in BIOS? I assume at that voltage you cant really stress test.
My reply to your post in another thread answers this very question = voltage identification LOL. :D It's basically luck of the draw - my CPU requires only 1.3375v for stock speed whereas yours requires 1.35v. This difference in voltage requirement is made clearer by my overclock of 3.75GHz which is rock solid at 1.425v; 3.9GHz may need only 1.48v for prime stability, and I can do various things such as SPi and web browsing at 4GHz with only 1.512v. But 3.9GHz 24/7 is a very nice achievement with your VID. :up:
milan03
12-16-2007, 04:45 PM
My reply to your post in another thread answers this very question = voltage identification LOL. :D It's basically luck of the draw - my CPU requires only 1.3375v for stock speed whereas yours requires 1.35v. This difference in voltage requirement is made clearer by my overclock of 3.75GHz which is rock solid at 1.425v; 3.9GHz may need only 1.48v for prime stability, and I can do various things such as SPi and web browsing at 4GHz with only 1.512v. But 3.9GHz 24/7 is a very nice achievement with your VID. :up:
oh I see. so VID in coretemp is indeed the required ID voltage for my cpu.
I'm considering lowering down to maybe 3.7-3.8 just to see whats teh lowest voltage/heat i can achieve on those speeds. I do need to much V for 3.9... 1.57V in bios. See what i get for slightly lower clocks.
VID is the core Voltage Identification, i.e. its decided by MSR values which tell the system the amps each core is being fed. It doesn't have anything else to do with VCore but it's part of the equation of how hot and power consuming your CPU is.
If you require 1.33/1.35V Vcore for stock speeds, I think that's quite bad or maybe mine was good in terms of headroom with temps/power. I used to mostly run mine at 400x8 or 400x7 and it only needed 1.20VCore BIOS for 2800MHz. That was the same in windows though and load was 1.189V. Mine was 1.35V VID.
itslogz
12-16-2007, 09:10 PM
Cool cool.
I can get my e6750 to 3.7ghz with 1.500 Bios set, stable. Pencil vdroop modded and voltage damper enabled....
3.8 puts the temps where I dont like them right now, because im waiting on my lapping kit to lap this cpu since when i lapped my old e6600 i dropped load temps 10C, so im waiting to lap to go to 3.8 since its gonna take over 1.525vcore more than likely...
milan03
12-16-2007, 09:56 PM
Cool cool.
I can get my e6750 to 3.7ghz with 1.500 Bios set, stable. Pencil vdroop modded and voltage damper enabled....
3.8 puts the temps where I dont like them right now, because im waiting on my lapping kit to lap this cpu since when i lapped my old e6600 i dropped load temps 10C, so im waiting to lap to go to 3.8 since its gonna take over 1.525vcore more than likely...
I basically can get my e6750 orthos stable at 3.7 at about 1.45In bios. then to get to 3.8 i need a lot more.. over 1.5. For 3.9ghz i need about 1.57vcore. The temps seem to be the best at 3.7 and 1.45Vcore. under 50C under stress test. Maybe that is where i should be keeping my cpu 24/7...
Have you tested max FSB on the chip/board?
I've just had that board and chip a few days earlier. The board had a big vdrop and vdroop.
milan03
12-17-2007, 07:44 AM
yeah its kinda big. .07ish
Spyrus
12-17-2007, 12:57 PM
Well not bad but you could do better @ cooling is a scythe mugen + papst
This is my 24/7 setup
http://www.imageshack.gr/files/c2pyrb6ihei5mgx9b8es.jpg (http://www.imageshack.gr/view.php?file=c2pyrb6ihei5mgx9b8es.jpg)
This is max post for now @1.55V
http://www.imageshack.gr/files/df05x6jc2a9mjbor0s8g.jpg (http://www.imageshack.gr/view.php?file=df05x6jc2a9mjbor0s8g.jpg)
Max 1M @ 1.55 V
http://www.imageshack.gr/files/s85ulh538urgx5koysek.jpg (http://www.imageshack.gr/view.php?file=s85ulh538urgx5koysek.jpg)
milan03
12-17-2007, 01:05 PM
I can SuperPi at over 4ghz, but thats not the point if its not 24/7 Orthos priority 9 small fft's stable. What am I gonna do with an unstable PC ?
I think your 24/7 setup could do better with thermalright u-120 extreme :)
LennyRhys
12-17-2007, 03:35 PM
Well not bad but you could do better @ cooling is a scythe mugen + papst
I find your post extremely unhelpful and pretty self-absorbed TBH. It's great that you can do 4.2GHz at 1.55v, but there's no need to belittle the efforts of a fellow overclocker whose CPU clearly has a higher voltage requirement than yours does. Nothing in your post is of any assistance to milan03 in his attempt to achieve a prime-stable 4GHz. :down:
milan03
12-17-2007, 04:33 PM
The funny thing is that at 3.6Ghz I need so much less voltage and therefore less heat than at lets say 3.8-3.9Ghz. Since my system is a HTPC at only 36C full load, I could totally turn all my case fans off and enjoy silent PC. It makes me think wether thats the optimal clock for my system... too bad I'm never satisfied :)
450FSB 3600 was my preferred setting too. ;) Very easy going and dead stable. A good abalnce of performance I reckon. Not many people running higher FSBs because the voltages and hence all the heat of every component will be high and unless you spend a lot on cooling, that's risky and always more stressful on your components.
We know your trying for 24/7 stability as in every day casual use. Would it really make that much of a perf/time/cost difference to you to run 3900 over 3600 or 3700?
LennyRhys
12-17-2007, 04:43 PM
That CPU cooler is simply awesome, and your CPU is great too if you can get 3.6GHz stable with so little voltage. So far I've found 3.6GHz to be Orthos stable at 1.387v, but with a little tinkering I reckon I could probably get it a bit lower. With the stock cooler the load temps are 62C. :D
As far as silent PCs go, I can turn off all five case fans and my PC is still audible - it's the stock CPU and GPU fans that make all the noise in my system lol.
BeastNotro
12-17-2007, 05:00 PM
Try for 4.0-4.2ghz.
Jodiuh
12-17-2007, 05:10 PM
Grrr...
E6400
1.350VID
3400Mhz
1.36v in OS
25C ambient
TRUE w/ SFLEX E
Orthos Blend = 58C load
I'm not even in a case! :( My TRUE's pipes have never felt even warm and even now they're almost cool. Small dab, large dab, both lapped, contact's fine...is my TRUE broke?
milan03
12-17-2007, 05:35 PM
That CPU cooler is simply awesome, and your CPU is great too if you can get 3.6GHz stable with so little voltage. So far I've found 3.6GHz to be Orthos stable at 1.387v, but with a little tinkering I reckon I could probably get it a bit lower. With the stock cooler the load temps are 62C. :D
As far as silent PCs go, I can turn off all five case fans and my PC is still audible - it's the stock CPU and GPU fans that make all the noise in my system lol.
hehe you know whats even better? my RAM will finally get to the 4-4-4-12 timings at 2.2V, while at higher clock i had to go with 5-5-5-18.
This could be a keeper :toast:
LennyRhys
12-17-2007, 05:35 PM
Try for 4.0-4.2ghz.
You obviously haven't paid any attention to the rest of the thread! :D
hehe you know whats even better? my RAM will finally get to the 4-4-4-12 timings at 2.2V, while at higher clock i had to go with 5-5-5-18.
This could be a keeper
I tried 4-4-4-12 but I think it's too much for my OCZ. It's rated at 4-4-4-15, and at 4-4-4-14 it gives me great SPi results. Highest I've had it so far is 2.15v, and if I'm brave I'll bump it to 2.2v should my system require further stabilizing with the FSB at 501MHz. :D And if I do ever manage to bench 3DMark06 at 4GHz and SPi at 4.1GHz or something like that, I may call it a day. All this stress on my system is worrying me. ^^
milan03
12-17-2007, 05:46 PM
You obviously haven't paid any attention to the rest of the thread! :D
I tried 4-4-4-12 but I think it's too much for my OCZ. It's rated at 4-4-4-15, and at 4-4-4-14 it gives me great SPi results. Highest I've had it so far is 2.15v, and if I'm brave I'll bump it to 2.2v should my system require further stabilizing with the FSB at 501MHz. :D And if I do ever manage to bench 3DMark06 at 4GHz and SPi at 4.1GHz or something like that, I may call it a day. All this stress on my system is worrying me. ^^
Well your chip seem to have incredible low voltage needs! If my chip was able to hit those high clocks stable... oh boy, i'd be keeping it up there at least untill every single person convices me that i don't really need that high clock for everyday HTPC use (which i really dont) hehee :shrug:
Spyrus
12-18-2007, 02:00 AM
I can SuperPi at over 4ghz, but thats not the point if its not 24/7 Orthos priority 9 small fft's stable. What am I gonna do with an unstable PC ?
I think your 24/7 setup could do better with thermalright u-120 extreme :)
Mine is orthos stable at 8x482 but even the voltage is only 1.4V the temperatures are high there. Of course with any other application i am fine.
The point of my post where that even someone can bench at that high with little vcore too he cannot have a cpu more than 3.6-3.7 for 24/7 for everyday. :up:
Spyrus
12-18-2007, 02:02 AM
I find your post extremely unhelpful and pretty self-absorbed TBH. It's great that you can do 4.2GHz at 1.55v, but there's no need to belittle the efforts of a fellow overclocker whose CPU clearly has a higher voltage requirement than yours does. Nothing in your post is of any assistance to milan03 in his attempt to achieve a prime-stable 4GHz. :down:
As i said in my previous post i doubt he can be prime-stable at 4GHz.
Even if he had a good cpu like mine.That is the point. But he could lower
the temperatures more if he can with air-case flow to achive at around 3.8 :up:
Spyrus
12-18-2007, 02:04 AM
450FSB 3600 was my preferred setting too. ;) Very easy going and dead stable. A good abalnce of performance I reckon. Not many people running higher FSBs because the voltages and hence all the heat of every component will be high and unless you spend a lot on cooling, that's risky and always more stressful on your components.
We know your trying for 24/7 stability as in every day casual use. Would it really make that much of a perf/time/cost difference to you to run 3900 over 3600 or 3700?
I aggree on that 3.6 is far fair for performance and not stressing the system.
Also i can have all fans to ~1000 rpm (120mm) so it is very silent :up:
Spyrus
12-18-2007, 02:07 AM
hehe you know whats even better? my RAM will finally get to the 4-4-4-12 timings at 2.2V, while at higher clock i had to go with 5-5-5-18.
This could be a keeper :toast:
It is better to use a divider 1:1.2 like me at 540 with 5-5-5-x is faster than
1:1@450@4-4-4-x.
You can test it yourself to see that
If your mobo can do 1:1.25 divider (mine does not) and your memory too
you can have even better results.
At that point you are stressing mch and less vdimm (but mobo need to be stressed) :D
mad_skills
12-18-2007, 03:21 AM
Well your chip seem to have incredible low voltage needs! If my chip was able to hit those high clocks stable... oh boy, i'd be keeping it up there at least untill every single person convices me that i don't really need that high clock for everyday HTPC use (which i really dont) hehee :shrug:
Just use 4GHz and above speeds for benchmarking and SPi and etc.
No need to stress your system that much. As for your cpu don't worry, i've seen E6750's here on the forum, that need 1.5V :D for 3.6GHz Orthos stability, so if you look it that way your cpu is very good. :)
As for me i need 1.37500V for 3.6gigs stable but for 4gigs i need 1.5V and a bit above so i dont bother, not even on TRUE120..to much vcore for 24/7 on air.
Poz i ziv bio.:up:
LennyRhys
12-18-2007, 03:57 AM
As for your cpu don't worry, i've seen E6750's here on the forum, that need 1.5V :D for 3.6GHz Orthos stability,
There is actually one that needs 1.59v for 3.6GHz :eek:
I'm very fortunate that my CPU can bench 4GHz with the stock cooler, and can do 3DMark06 at 3.75GHz also with the stock cooler. I overclock strictly for benching; for general use (including games) there is simply no need to have my system overclocked. I do like the idea of a generally faster PC, but only when I get my Zalman will I contemplate a 24/7 CPU clock over 3.6GHz. :yepp:
mad_skills
12-18-2007, 05:00 AM
1.59V ? Now that's uber bad luck.:down:
Rob94hawk
12-18-2007, 05:29 AM
I like this thread. Can't wait to burn the E6750 like you guys soon.
milan03
12-18-2007, 07:37 AM
As i said in my previous post i doubt he can be prime-stable at 4GHz.
Even if he had a good cpu like mine.That is the point. But he could lower
the temperatures more if he can with air-case flow to achive at around 3.8 :up:
Bro, The whole post is about my chip being able to Prime stable at 3.9 although with 1.57V in bios. Check out the first screenshot. Not sure what is the problem...:confused:
milan03
12-18-2007, 07:40 AM
-EDIT- installed it.
Spyrus
12-19-2007, 01:15 AM
Maybe it is the mobo.
Raise vmch to 1.56 volt at least.
Keep temperatures ( seeing them from coretemp only not uguru) below 70 on load.
Raise vtt if needed.
Be sure that your memory can handle 500+ (put in 5-5-5-x)
Inteleron
12-19-2007, 06:50 PM
Nice clocks guys. Im interested in picking up a 6750 + MSI P35-NEO2-FR to replace my 6300 & Biostar mobo in my #1 rig. I was wondering what stepping/week to look for? do most 6750's do 3.6 on air out of the box?
Spyrus
12-20-2007, 02:07 AM
You have already a cpu in 3.5 so do not bother.
Go for quad instead.
mad_skills
12-20-2007, 02:36 AM
Nice clocks guys. Im interested in picking up a 6750 + MSI P35-NEO2-FR to replace my 6300 & Biostar mobo in my #1 rig. I was wondering what stepping/week to look for? do most 6750's do 3.6 on air out of the box?
Look for any week in october. E6750's from October weeks overclock much better than the ones before. All E6750's are G0 stepping.
I don't think it's smart to do 3.6GHz on box cooler, firstly too much heat, secondly you can get a nice cooler for a not too great amount of money.
And i think that that mobo is not such a good choise for o/c, not bad but for a little bit much cash, you can get a DFI Blood Iron P35-T2RL.
LennyRhys
12-20-2007, 03:41 AM
If I was looking for a new CPU I'd still choose the E6750 over the Q6600. Four cores are more expensive, harder to overclock, require more efficient cooling, and (most importantly) aren't necessary, even though some people like to think so. Dual core FTW. :D
mad_skills
12-20-2007, 04:20 AM
I think quad core is a good choise but not the Q6600, i would wait for Q9450, the cheapest one with 12Mb L2, uber best-buy.
Inteleron
12-20-2007, 10:16 AM
yea guys, I need to upgrade my motherboard because although my Biostar is a great clocker, it doesn't have raid, and I need more hdd space so i was going to raid 2 Seagate 250's. While i was at it i figured Id pawn my E6300 and grab a 6750 which has a higher multi, double the cache, and maybe a nice 3.6-3.7GHz oc under my Zalman 9700.
milan03
12-20-2007, 02:58 PM
Go with E6750 man. Mine was packed 11/20/07 made in costa rica.
Batch# 3741B003
Good luck!
Rob94hawk
12-23-2007, 09:00 PM
Go with E6750 man. Mine was packed 11/20/07 made in costa rica.
Batch# 3741B003
Good luck!
Is the batch # anywhere else besides on the heat spreader? On the box maybe? I just put the Thermalright 120 extreme on it today and forgot to snap a photo of it.:( What are the best batches?
Do you guys think my X38/DDR3 combo will help me get my E6750 to a stable 4Ghz? So far I'm only at 3.2Ghz(400x8)/DDR3 1600 with stock cpu voltage.
milan03
12-23-2007, 11:07 PM
Is the batch # anywhere else besides on the heat spreader? On the box maybe? I just put the Thermalright 120 extreme on it today and forgot to snap a photo of it.:( What are the best batches?
Do you guys think my X38/DDR3 combo will help me get my E6750 to a stable 4Ghz? So far I'm only at 3.2Ghz(400x8)/DDR3 1600 with stock cpu voltage.
Batch# is on the box. As for the 4ghz OC, why dont you try and see. Every CPU is different, and only you will know how far can you push your cpu.
Cheers.
running mine @ whats in my sig check pic
milan03
12-24-2007, 09:45 AM
running mine @ whats in my sig check pic
Looks good there! We have the same cpu and ram/timings at 450mhz fsb. Just not sure why your vcore is so high for 3.6Ghz! Thats over 1.5V in Bios!!!
My chip needs only 1.39V in bios at 3.6Ghz for long orthos session. Try lowering down that voltage and see if you can find the lowest stable voltage at that clock.
CHeers!
mad_skills
12-24-2007, 09:48 AM
I've got even better than milan's. I have 3.6GHz stability @ 1.37500V. :D
@BTK You should try and lower the vcore, that's a bit high.
Spyrus
12-24-2007, 10:06 AM
And i have better than mad_skills ...
I have 3.6GHz stability @ 1.300V :rofl:
To be serious now my batch is L724B019.
There is one other guy with this batch at 3.75 GHz.
I think it is super.
But most important is the wafer code.
It is at the bottom of cpu written on cpu not the hsf.
Mine is N472500
The last two digit show that i have the first of the wafer and the better one.
So combine week+wafer and you have an astonished 4.2GHz benchmarkable cpu
at only 1.58Volts like mine. :up:
milan03
12-24-2007, 10:09 AM
And i have better than mad_skills ...
I have 3.6GHz stability @ 1.300V :rofl:
To be serious now my batch is L724B019.
There is one other guy with this batch at 3.75 GHz.
I think it is super.
But most important is the wafer code.
It is at the bottom of cpu written on cpu not the hsf.
Mine is N472500
The last two digit show that i have the first of the wafer and the better one.
So combine week+wafer and you have an astonished 4.2GHz benchmarkable cpu
at only 1.58Volts like mine. :up:
Bravo for you! :up:
mad_skills
12-24-2007, 10:10 AM
That's one good cpu you got there.:up:
Honestly i didn't try any less, but i got 4GHz prime stable at around 1.5V, a bit higher though, that's an ok score i believe..but when i see E6750 cpu's that need 1.55V or similar for 3.6GHz stability i feel very lucky...:D
milan03
12-24-2007, 10:13 AM
That's one good cpu you got there.:up:
Honestly i didn't try any less, but i got 4GHz prime stable at around 1.5V, a bit higher though, that's an ok score i believe..but when i see E6750 cpu's that need 1.55V or similar for 3.6GHz stability i feel very lucky...:D
Totally! Plus, I just read something on this forum about the new 45nm chips dying so easy... made me appreciate my low volts. I'm gonna chill at 3.6Ghz for a while :)
Spyrus
12-24-2007, 11:13 AM
Well latest batches are also good for Quad 6600.
I was in a pc store this morning and i was discussing with the
kid there.He said that this month's quad can easily do 4.2 on air (benchable)!
I think Intel has put something of Penryn/Wolfdale in there.
Spyrus
12-24-2007, 11:20 AM
Totally! Plus, I just read something on this forum about the new 45nm chips dying so easy... made me appreciate my low volts. I'm gonna chill at 3.6Ghz for a while :)
Is there a link?
I am really curious about that :confused:
I've got even better than milan's. I have 3.6GHz stability @ 1.37500V. :D
@BTK You should try and lower the vcore, that's a bit high.
thats idle
actual load is 1.440V
http://img220.imageshack.us/img220/8462/smallfttbp6.jpg
anything lower core #2 it fails
my chip is a good oc'ing chip as it hits 4 ghz easily but o actually be stable it likes voltage
3.5 ghz is stable @ 1.392V load while 3.6 takes 1.440V load
not sure if i got a bad chip did i?
milan03
12-24-2007, 12:18 PM
Is there a link?
I am really curious about that :confused:
http://www.xtremesystems.org/forums/showthread.php?t=170358
@ BTK: My chip needs no Volt increase at anything below 3.6Ghz and I'm talking 5+ hours prime stable. at 3.6Ghz I have it up a notch 1.37V but that voltage isnt enough for 5+ hours of prime. Failed after close to 6 hours. At 1.39 its prime stable 24 hours. The funny thing is that at 3.8Ghz I need quite a lot volts about 1.49V, at 3.9Ghz I need 1.57V and at 4ghz(unstable) I need about 1.6V.
Imho, the best for the system is to find the sweet spot where you get the highest overclock for the least Volts alteration, in my case thats a little over 3.6Ghz. I keep it at 3.6Ghz. Also worth mentioning that at 3.8Ghz, I need to start messing with the NorthBridge settings as well as with the VTT. All that ads unnecessary heat.
In your case I'd keep it at 3.5Ghz for everyday use since the difference between 35 and 3.6 is so little and the vcore jump on your chip is quite noticeable.
Hope this helps ;)
mad_skills
12-24-2007, 12:35 PM
You haven't got a bad one, just an average..it's ok but the the best.;)
even with the 1.440c load i dont worry about it because the tuniq keeps the temps very good
58C prime95 small ftt load
vs 56c 3.5
tuniq does good with hig vcore
mad_skills
12-24-2007, 01:47 PM
Yeah, Tuniq and TR do great job at high vcore.
As long it's under 60C and it is, you are doing a nice job.;)