E6750 Retail - Some results Air Cooled

[MikalCarbine]

Thats strange, and you have the same board as me
http://www.xtremesystems.org/forums/...or/menupop.gif
http://www.xtremesystems.org/forums/...or/menupop.gif
Currently I'm priming at 3.68Ghz with 1.46 in BIOS, uguru reads ~1.41 but I keep getting random BSODs so I am playing with GTL/VTT and MCH voltages

I guess I got really lucky with this chip? I am hoping for 3.8 with decent voltages

[Rab]

got one of these this week, stepping L719B030
sitting @ 3.4 with 1.41 v, temps idle 36-39c full load 61-63c, this on stock cooler. (ultra soon) For a £120chip very happy with this, hoping can do a lot more with better cooling

[Drizzthsz]

Hi, this is my first post. So, what are the experinces so far? If i will have the chance to check the Fpo of the processor, for which numbers should I look for? Would you suggest an L719B rather than an A?
Thx.

[Rab]

upped to 3.6ghz, seems stable
what kind of temp drop will the ultra give me over stock cooler, -10c more ?

[Drizzthsz]

upped to 3.6ghz, seems stable
what kind of temp drop will the ultra give me over stock cooler, -10c more ?

Very nice results with stock cooling, I hope to get one that overclocks this nice.

[kniwor]

The results seem good so far, most of u guys have not mentioned what type of air cooling u're using, also, it would be great if someone took the trouble of tabulating the results. Going to get an E6750, prolly tomorrow, will post the results soon.

Temperatures still are confusing, no way in hell can the temps be lower than room temperatures on air, it's just against thermodynamics, even water would not give lower temps than ambient, unless water itself is cooled by external source.

[MikalCarbine]

I think I asked this in some thread earlier, but what are safe volts for these chips under water for 24/7? ~1.55?

[lordkelmain]

These look quite good. Can't wait to get mine booted up.

- lk

[DerekT]

Motherboard: ASUS P5k Vanilla ||| BIOS: 0603

CPU: E6750 ||| Batch#: L722A494 ||| Core: 3600Mhz ||| Mult: 8x ||| Bus: 450Mhz ||| FSB: 1800Mhz ||| Voltage: 1.48v

Memory: Crucial Ballistix Tracer 8500 2GB (2x 1GB) ||| Ratio: 5:4 ||| Frequency: 1125Mhz ||| Timings: 5-5-5-15 ||| Voltage: 2.15v

Cooling: Sunbeam Tuniq Tower 120 @ 47C full Orthos (11hours) Load temps.

I have no doubt that these revisioned CPU's can handle 1.6V on air. My next step is 3.8Ghz.


I have two E6600 (L628B, L629B) and both can also make 3.6Ghz easy but the temps disallow running them higher. They get close to 70C at 3.6Ghz. They need far less voltage but run much hotter.

Here's an interesting point to note. I am using an ASUS P5K Vanilla (Bearlake chipset) based board and the latest BIOS (0603) which allows a choice of Strapping (200 - 800Mhz, 266 - 1066Mhz, and 333 - 1333Mhz). When I chose the 333 and clocked it up to 3.6Ghz I chose a lower (900Mhz) DRAM throughput and stability was quite tough to gain. When I dropped the Strap to 266 - 1066 and clocked it up to 3.6Ghz, I had 1125MHz on the RAM and stability was far easier to maintain. I am now at 1125 with 5-5-5-12. The Crucial Ballistix 1066 D9GMH has no problems up to 1285 @ 5-5-5-15. Something tells me that DDR3 will allow us to gain some great clocks on this revisioned CPU.

[Yori1989]

Rofl, my cpu @ 3.3ghz with 1.35vcore is 56degrees celcius @ load.

Somethings guess 3 times, i think my ISH isnt flat :/

[msgclb]

Rofl, my cpu @ 3.3ghz with 1.35vcore is 56degrees celcius @ load.

Somethings guess 3 times, i think my ISH isnt flat :/

The highest load CPU Temp that I've seen was 38c @3500MHz - 500FSB x7 DDR2-1000 when I temporarily had vcore at 1.6v. Using 1.35vcore the load temp stays around 25c-26c. My ambient temp is ~24c-25c. I'd bet the real temp is +15c. This is on water.

It almost looks like these G0 CPUs could run without a HSF.

I wouldn't bet on not being flat as causing your high temps. I'd bet on it not being seated properly.

[Yori1989]

The highest load CPU Temp that I've seen was 38c @3500MHz - 500FSB x7 DDR2-1000 when I temporarily had vcore at 1.6v. Using 1.35vcore the load temp stays around 25c-26c. My ambient temp is ~24c-25c. I'd bet the real temp is +15c. This is on water.

It almost looks like these G0 CPUs could run without a HSF.

I wouldn't bet on not being flat as causing your high temps. I'd bet on it not being seated properly.

reapplied, load temp is 54 degrees celsius :/(per core)

I still think its the IHS

[Mourngrym]

E6750 L722A421, Tuniq Tower 120, Asus P5K-E, 2GB Corsair 6400C4, 8800GTX, Corsair 520W, Antec P182.

Now running stable at 8x425=3400 Mhz @ 1.475 Vcore-BIOS (1.448 in CPU-Z, 1.440 load)

8x440=3520 Mhz was not Orthos stable at 1.5 Vcore-BIOS (1.472 CPU-Z). Decided I didn't want to go higher on the voltage. But I did try 8x450 @ 1.55 Vcore-BIOS just for fun, but was unstable.

These E6750s seem a mixed bag indeed.

[Mourngrym]

Cooling: Sunbeam Tuniq Tower 120 @ 47C full Orthos (11hours) Load temps.

I'm guessing that you haven't taken the different temp readings for E6750 into account. Disable PECI in BIOS, load up SpeedFan and add +15C to the core temps. I suspect you will end up around 62C. SpeedFan also shows the "CPU" temp which is the sensor between the two cores. Compare with that and you will see if the reading is reasonable. My temps are about 63-64 core and 61-62 CPU with full Orthos, at 3400 Mhz, 1.475 Vcore-BIOS and Tuniq @ 1500RPM. Idles at about 37C. No one knows 100% sure how to read the temps yet though, but these seem reasonable to me.

[RPGWiZaRD]

Man all these negative results about E6750 makes me only more nervous that I ordered one on friday. My goal is 3.6GHz+ stable on air, if I don't reach that I'll prolly sell it. xD The results varies so much, some users have reached 3.4GHz even on stock volts, others need close to 1.5v for that lol and I've seen huge variance even for exact same stepping. All that makes me wonder if the results aren't highly dependant on the mobo you have...

[slayer_warrior]

(e6750 go batch L722a494)

i did 3.4 ghz on stock 1.35v but after that i had to up the volts 2 times for ever 5 mhz get stable in orth stop at 3.6 1.43 but crash after couple hours got it stable at 1.45 underload 1.42v 6 hours orth stable before i stop orth. but the temp under load 39/40c in tat,core temp .94, speedfan these temp can not be right. idle they are 9/10c i live in the south! i'm very happy with this chip!

coming from amd 4000 single core stock to intel was very good upgrade very very happy >:P. runs my games alot better

e6750 go (1.45v 450x8 3.6ghz)
ga p35-ds3p stock volts
xfx7800 gtx oc 490/1.2ghz
g.skill 6400 hz 2.1v 900mhz 4-4-4-12 with 80 mm fan cooling
wd 16mb 250 gig sata2 hdd
sony dvd burner
cheap dvd,cd rom need one was short on cash lol.
ocz powersteam 520
5 80mm case fans

[DerekT]

I'm guessing that you haven't taken the different temp readings for E6750 into account. Disable PECI in BIOS, load up SpeedFan and add +15C to the core temps. I suspect you will end up around 62C. SpeedFan also shows the "CPU" temp which is the sensor between the two cores. Compare with that and you will see if the reading is reasonable. My temps are about 63-64 core and 61-62 CPU with full Orthos, at 3400 Mhz, 1.475 Vcore-BIOS and Tuniq @ 1500RPM. Idles at about 37C. No one knows 100% sure how to read the temps yet though, but these seem reasonable to me.

The only difference is that when I add 15+ C to the values, they increase.

LOL

You say nobody is quite sure yet, then you ask to arbitrarily increase the values to find values that are close to the prior stepping/revision.

My answer?

Uh, no, I don't have to do that. I can easily tell by just testing the differential between a 6600 L630A and this new chip. I installed stock coolers, then set the values both to 8 X and ramp up the FSB to 450. Then I tested the warmth outputted with a digital thermal readout. The L630A is definitely warmer than this new one. For my two processors tested, it is greater than 10C and I can only speak to my two tested processors. Time will tell, but your method makes no real scientific sense.

Add 15C arbitrarily? That makes no sense unless you knew for sure the values of differentiation. You do not so what you say makes no sense and no sense = nonsense.

That is not to say that there is not a differential. Just that your arbitrarily chosen offset is not scientific or proven whatsoever.

[kniwor]

Was just testing my E6750 today, the processor seem pretty decent to me, the more voltage I give, the higher it goes, but the problem is my temperature seems to be rising a bit too much, it's stable at 3600Mhx@1.5V in BIOS of ASUS-P5B-Deluxe.

How do I know the actual Vcore, CPUZ gives absurd readings.

What is the tjunction for the E6750 processor. My delta to Tjunction goes to as low as 20c on both cores while running it 3600Mhz at 1.5V.

It is stable at 3200Mhz at 1.3125V, I havent yet tested it at higher than 1.5V in BIOS as I was too concerned about the temperatures, will apply AS5 later and then see if that helps, I'm on stock cooling, and I cant post the screenies right now due to problems in my internet connection.

[DerekT]

Was just testing my E6750 today, the processor seem pretty decent to me, the more voltage I give, the higher it goes, but the problem is my temperature seems to be rising a bit too much, it's stable at 3600Mhx@1.5V in BIOS of ASUS-P5B-Deluxe.

That's where I stand with my setup, and yes, it desires 1.5V as less does not give me long term stability, and I want long term proof for purposes of folding.

How do I know the actual Vcore, CPUZ gives absurd readings.

What is the tjunction for the E6750 processor. My delta to Tjunction goes to as low as 20c on both cores while running it 3600Mhz at 1.5V.

You can download Everest which will give you a more true voltage reading or you can use ASUS Probe II which will as well. With regard to the CPU idle thermals, it would seem that a new offset needs to be created for this new revision. I had personally thought that with the release of the Intel digital thermal sensors embedded singularly in the core, this would be a thing of the past. It's rather confusing as to why the DTS does not give a true reading. It's also quite nonsensical that the CPU reading would be of a higher thermal than the core readings. This is contrary to simple electronic physics.

Have you downloaded the latest BIOS (P5B Deluxe BIOS version 1215)?

http://support.asus.com/download/dow...Language=en-us

EDIT:

I'm on stock cooling.

I never use stock cooling, and I would think that you might well find thermal and other issues with stock cooling. Also, even though P5B has allowed a 1333FSB processor, there might well be upcoming BIOS fixes that correct and offset the temperatures to bring them more in line at the low level and possibly at the top level as well.

I advise you to move to better cooling. 3.6Ghz and 1.5V is not a good spot for the Intel Stock Heatsink.

[Mourngrym]

The only difference is that when I add 15+ C to the values, they increase.

LOL

You say nobody is quite sure yet, then you ask to arbitrarily increase the values to find values that are close to the prior stepping/revision.

My answer?

Uh, no, I don't have to do that. I can easily tell by just testing the differential between a 6600 L630A and this new chip. I installed stock coolers, then set the values both to 8 X and ramp up the FSB to 450. Then I tested the warmth outputted with a digital thermal readout. The L630A is definitely warmer than this new one. For my two processors tested, it is greater than 10C and I can only speak to my two tested processors. Time will tell, but your method makes no real scientific sense.

Add 15C arbitrarily? That makes no sense unless you knew for sure the values of differentiation. You do not so what you say makes no sense and no sense = nonsense.

That is not to say that there is not a differential. Just that your arbitrarily chosen offset is not scientific or proven whatsoever.

True, no scientific proof or anything. I just wanted to point out that your temps *may* be on the low side. And of course you don't *have* to do anything, it was just some information for you that you may judge as you see fit. As for it being arbitrary, well yes, but it currently seems like the most reasonable educated guess that I have found. This would be based on the assumtion that Intel has chosen Tjunction = 100C for these CPUs instead of 85C. Also, when I compare the core temps reported by SpeedFan with the temp reported for "CPU", it seems reasonable that the cores are a few degrees higher. I find it unlikely that the core temps would be lower than what the "CPU" sensor shows. I personally would rather be pessimistic on temperature than optimistic until we know. Hopefully Intel will reveal this soon.

[DerekT]

True, no scientific proof or anything. I just wanted to point out that your temps *may* be on the low side. And of course you don't *have* to do anything, it was just some information for you that you may judge as you see fit. As for it being arbitrary, well yes, but it currently seems like the most reasonable educated guess that I have found. This would be based on the assumtion that Intel has chosen Tjunction = 100C for these CPUs instead of 85C. Also, when I compare the core temps reported by SpeedFan with the temp reported for "CPU", it seems reasonable that the cores are a few degrees higher. I find it unlikely that the core temps would be lower than what the "CPU" sensor shows. I personally would rather be pessimistic on temperature than optimistic until we know. Hopefully Intel will reveal this soon.

I agree that the temps *may* well be inaccurate. I believe that they are, however I do not agree with the arbitrarily chosen 15+ C addition. Regardless of a T-Junction of 100 for the L stepped revision, the A and B stepped are still at 85C and it is more likely that these G0's are not of those inferior fabricated processors.

You stated that I haven't taken the temperature differentiation into account. You need to say " possible" in that statement in my view. That statement presupposes that it is not a guess and that the information is known. It is not, and your arbitrarily chosen 15+C increase is not at all (at this time) indicative of reality.

So, even though I agree with you that thermal differentiation is very possible and likely, I have to disagree with your choice of a 100 - 85 = 15C+ quantification. Simple thermal testing has shown me that there is in fact a fair difference of over 10C between the L630A (E6600) and the L722A(E6750) when both cpu's are running 450MHz FSB and an 8X multi. The fact of the FAB (LxxxAxxx) shows that it is somewhat doubtful that the T-Junction has changed although it is possible. Certainly my thermals show the opposite, both on internal core DTS and applied (IHS) thermal sensor. Thus I disagree with your statement of "Educated Guess".

[HeavyH20]

The only thing about the E6750 that hurts is the 8x multi. My 680i board limits the OC to 474 (1896) FSB to net 3792 MHz. That is with 1.52 VCore.

[Mourngrym]

I agree that the temps *may* well be inaccurate. I believe that they are, however I do not agree with the arbitrarily chosen 15+ C addition. Regardless of a T-Junction of 100 for the L stepped revision, the A and B stepped are still at 85C and it is more likely that these G0's are not of those inferior fabricated processors.

You stated that I haven't taken the temperature differentiation into account. You need to say " possible" in that statement in my view. That statement presupposes that it is not a guess and that the information is known. It is not, and your arbitrarily chosen 15+C increase is not at all (at this time) indicative of reality.

So, even though I agree with you that thermal differentiation is very possible and likely, I have to disagree with your choice of a 100 - 85 = 15C+ quantification. Simple thermal testing has shown me that there is in fact a fair difference of over 10C between the L630A (E6600) and the L722A(E6750) when both cpu's are running 450MHz FSB and an 8X multi. The fact of the FAB (LxxxAxxx) shows that it is somewhat doubtful that the T-Junction has changed although it is possible. Certainly my thermals show the opposite, both on internal core DTS and applied (IHS) thermal sensor. Thus I disagree with your statement of "Educated Guess".

The reason why I replied was because I see many incorrect (too low) temps posted on various forums. It's important to note that they might very well be higher for those who do not know (so they don't damage something). I might not have used the correct wording, my apologies for that. Not everyone has english as their native tongue.

Peace.

[DerekT]

The reason why I replied was because I see many incorrect (too low) temps posted on various forums.

Indeed, and I agree with you on this view. I am certain that these CPU's do in fact run cooler. How much is the point of contention at this time.

I have two E6600 (L628B and L629B) that run far warmer than the E6750 (L722A). Even my cooler E6600 (L630A) runs a good bit warmer than the E6750 and this is with an IHS applied thermal sensor. ~15C for the DTS and ~9C with the L62xB's and ~11C for the DTS and ~6C with the L630A. I had to drop the FSB of the L62xB's to 400 X 9 rather than 450 X 8 as they ran ~60C at 400 X 9 vs 72C @ 450 X 8. There's nowhere near that heat output on the G0 L722A. Anyone in the Lower Fraser Valley of BC can buy these E6600's off of me as I am moving entirely to the G0's.

[alex17 GTX]

Guys. Have you tested if the cpu:dram multipliers works? I mean : can you use something else than 1:1 on 965p motherboards?
I'm asking this becouse somenone tested an e6750 on a dfi dark and other ratio than 1:1 doesn't works.