Whats the furthest you've been able to overclock your Core 2 without adjusting voltage?

[ShadowZERO]

We all know pushing the core voltage of a CPU is the easiest way to increase stability while overclocking. Coming from the perspective of an electrician however, adjusting the voltage past manufacturer specs would be the most likely way to cause damage to the processor transistors over long-term usage.

I splurged on my system. I bought 7-7-7-20 1600mhz Corsair memory to go along with my 3.0 core 2 duo and Asus P5K3 Deluxe. Using the N.O.S. BIOS auto-overclock, I can overclock my 3.0 to 3.45, and my FSB strap automatically adjusts my memory speed to 1536mhz(768 doubled) to go along with my quad-pumped 384mhz(upped from the 333 default) Core 2 FSB. My FSB and memory frequency remain completely synchronous, resulting in the most efficient performance for communication between the CPU and memory.

While trying to push to 3.6 1600@1600, I experienced reboots while in Windows under heavy load, which I attributed to the overclock. I'm fairly positive I could get it running stable at that or beyond with a small voltage increase. My "burn-in" test for stability is running Hellgate:London on max settings while batch converting audio files using dB Power Amp at the same time.

My system is fast enough(is there such a thing, heh?) for me to consider adjusting the voltage at this point. I'm curious to see what others have been able to achieve when overclocking a Core 2 without adjusting the voltage.

 

[Syzygies]

The right question to ask is where the curves cross: Find the minimum voltage at each overclock, and report when that crosses the stock voltage flat line.

 

[Tweakin]

I can push my e6750 up to 3.3GHz from 2.66 on stock vcore. This is Prime and Orthos stable for 48+ hours. Ran my system here for 2 weeks doing seti 24/7.

 

[DSF]

My E4500 goes to 2.93 while below stock voltage, but going over 3.0 GHz requires a voltage bump. Maybe I could push it a little more, but I don't really have any reason to.

 

[demiurge3141]

my e2140 does 440x8=3.52 on stock volts.

 

[Drsignguy]

Originally posted by: demiurge3141
my e2140 does 440x8=3.52 on stock volts.

WOW! What board do you have that chip on? And, what's your cooling situation?

 

[demiurge3141]

Originally posted by: Drsignguy

Originally posted by: demiurge3141
my e2140 does 440x8=3.52 on stock volts.

WOW! What board do you have that chip on? And, what's your cooling situation?

ip35-e with freezer 7 pro

 

[harpoon84]

My E4400 (stock 2GHz) does about 3GHz at stock volts, though I haven't done extensive stability testing as its not my 24/7 speed.

 

[bryanW1995]

Originally posted by: demiurge3141

Originally posted by: Drsignguy

Originally posted by: demiurge3141
my e2140 does 440x8=3.52 on stock volts.

WOW! What board do you have that chip on? And, what's your cooling situation?

ip35-e with freezer 7 pro

I'd love to see a 12 hour prime stable screenshot of that...

 

[Drsignguy]

Place your bets!

 

[ShadowZERO]

Originally posted by: Tweakin
I can push my e6750 up to 3.3GHz from 2.66 on stock vcore. This is Prime and Orthos stable for 48+ hours. Ran my system here for 2 weeks doing seti 24/7.

This seems pretty impressive to me, overclocking more than 600mhz with no vcore adjustment. Nice job!

my e2140 does 440x8=3.52 on stock volts.

If i recall, E2xx processors, while sold as Core 2s, are actually using the latest P4 core. I won't go into the downsides of the NetBurst architecture, but the upside was always that they are easy to overclock if you can keep them cool.

On that note, I'm still using the stock cooler that came with my E6850. I'm wondering if upgrading to an aftermarket cooler would get me a better overclock without needing to adjust the vcore, just due to heat. I remember reading somewhere that the heatpipe-towers with the side blowing fan are usually the best aftermarket CPU coolers.

I'm holding off on installing any new progs until I complete a fresh install with an XP SP3 slipstream disc. After that, I plan to run Prime95 for a day and log the temps with the ASUS tool. That way, I can report how hot the CPU is getting to get a better idea if the CPU temp could be causing the instability at 3.6 rather than a lack of voltage.

Comments and suggestions are welcome, especially if anyone has an idea of whether or not overheating is the issue for my overclock instability rather than the vcore. Hopefully I'll have some temps to add once I get my system re-configured.

Any comments or suggestions are welcome!

 

[harpoon84]

Originally posted by: ShadowZERO
If i recall, E2xx processors, while sold as Core 2s, are actually using the latest P4 core. I won't go into the downsides of the NetBurst architecture, but the upside was always that they are easy to overclock if you can keep them cool.

Not at all, the E21x0 chips are Core 2 based, just with a smaller cache.

 

[Idontcare]

Originally posted by: Syzygies
The right question to ask is where the curves cross: Find the minimum voltage at each overclock, and report when that crosses the stock voltage flat line.

http://i272.photobucket.com/al...ableVcoreversusGHz.jpg

Kind of like this. My VID for this particular QX6700 is 1.35V. So I can clock it to 3.5GHz on stock voltage.

This is with vaporphase cooled mind you (and a B3 quad), but it fits perfectly within the OP's context in that running a CPU at stock voltage and -30C temps is likely to improve lifespan and not degrade it.

 

[OCGuy]

To be honest, I dont trust alot of people who say "I got to 'X' on stock voltage." What many people dont realize is that even though they may have never touched a Vcore setting in BIOS, it is set on auto. (Im obviously not talking about seasoned overclockers.)

 

[Idontcare]

Originally posted by: Ocguy31
To be honest, I dont trust alot of people who say "I got to 'X' on stock voltage." What many people dont realize is that even though they may have never touched a Vcore setting in BIOS, it is set on auto. (Im obviously not talking about seasoned overclockers.)

I considered myself a seasoned overclocker (keep the snickering to yerself ) and I too fell into this trap when I upgraded from i680 mobo's to a P35 DS3L.

I was thinking "woo-hoo, 3.3GHz on stock voltage on this Q6600, who-rah!" Then I realized that Gigabyte programmed their DS3L's BIOS to allow overvolting above and beyond the CPU's VID so long as the auto Vcore did not exceed Intel's max specified VID (which was 1.35V for a Q6600 at the time, now its 1.50V).

So I agree with you 100%...you can't say you are running at X GHz at "stock volts" unless you actually went into the BIOS and manually set the Vcore to the CPU's listed VID from Coretemp.

 

[Tweakin]

Originally posted by: Ocguy31
To be honest, I dont trust alot of people who say "I got to 'X' on stock voltage." What many people dont realize is that even though they may have never touched a Vcore setting in BIOS, it is set on auto. (Im obviously not talking about seasoned overclockers.)

After setting the default vcore within the bios, I was able to push my e6750 up to 3.3GHz from 2.66 on stock vcore. This is Prime and Orthos stable for 48+ hours. Ran my system here for 2 weeks doing seti 24/7.

 

[Syzygies]

Originally posted by: Idontcare
So I agree with you 100%...you can't say you are running at X GHz at "stock volts" unless you actually went into the BIOS and manually set the Vcore to the CPU's listed VID from Coretemp.

There's no way I'm paying Microsoft for a copy of Windows just to run Coretemp (I'm on Linux, and yes, the in-OS overclocking tools are more primitive). but my EP35-DS3P F2 BIOS reports a cpu voltate of 1.27500 when left on auto for any sufficiently low cpu frequency. I never leave anything on auto (pretty much every auto setting seemed to work against me) but I do look at its choices before making mine. E.g. make my first guess at tRD after seeing its guess.

So your DS3L BIOS doesn't report its auto choices to you? Did you update it? Is it wrong? What do you think my 1.27500 means?

 

[Idontcare]

Originally posted by: Syzygies
So your DS3L BIOS doesn't report its auto choices to you? Did you update it? Is it wrong? What do you think my 1.27500 means?

Nope, yep, not necessarily (its an intentional feature), I bet your 1.2750 is your VID (as set by Intel at the factory).

 

[chriskwarren]

Q6600 at 3.0ghz stock volts (set manually as to avoid auto volts increase, everest confirms this). Each core gets to about low-mid 40c running 4 instances of prime (can't recall exact temps). I think I can push it further no prob though but I'm ascared

 

[happy medium]

Originally posted by: demiurge3141
my e2140 does 440x8=3.52 on stock volts.

I'll give you 100.00$ for it.

 

[Zapper48]

On a DFI 965 Dark,my E8400 would do 3.6ghz on 1.15v
My IP35 Pro won't go that low on default 1.250.

 

[demiurge3141]

I think I gave a 21 hour screenshot before:

http://i260.photobucket.com/al...jihe3141/2140stock.jpg

and here is it failing at a higher vcore due to heat:

http://i260.photobucket.com/al...ii21/jihe3141/fail.jpg

 

[honolululu]

I'm can get to 3.4GHz (8x425) on my 6750 with 1.35V.

Actually set to Normal in BIOS because I find EIST/CIE voltage regulation works better.
I jump all the way down to 1.168V from 1.312V on the multi drop. Seems to always stay above 1.2V when I set vcore to 1.35V in BIOS.

 

[PG]

Originally posted by: Idontcare

Originally posted by: Ocguy31
To be honest, I dont trust alot of people who say "I got to 'X' on stock voltage." What many people dont realize is that even though they may have never touched a Vcore setting in BIOS, it is set on auto. (Im obviously not talking about seasoned overclockers.)

I considered myself a seasoned overclocker (keep the snickering to yerself ) and I too fell into this trap when I upgraded from i680 mobo's to a P35 DS3L.

I was thinking "woo-hoo, 3.3GHz on stock voltage on this Q6600, who-rah!" Then I realized that Gigabyte programmed their DS3L's BIOS to allow overvolting above and beyond the CPU's VID so long as the auto Vcore did not exceed Intel's max specified VID (which was 1.35V for a Q6600 at the time, now its 1.50V).

So I agree with you 100%...you can't say you are running at X GHz at "stock volts" unless you actually went into the BIOS and manually set the Vcore to the CPU's listed VID from Coretemp.

The DS3L has some funny quirks that take some time to figure out. I have one for myself and built a system for a friend with another one.
I've messed around with the settings for hours before I finally figured out the best way to set these up.

If you leave the main voltage setting that controls everything at Auto, then yes, the board goes a bit nuts on the Vcore. For my E6300 and my friends E2160 the DS3L set the Vcore to about 1.39 which is not even remotely necessary.

You can set the voltage manually, but then you have to experiment before you figure out the best balance between having enough for stability, but you don't want to go too high because the temps get out of control.

Here is another way, and the best way to set these boards up:
1. Set the main voltage to Manual and not Auto.
2. Set all voltages to Normal, especially Vcore.
3. There is another screen in the bios with EIST and C1E. Enable both of those.
4. Save and exit.

Now download CPU-Z, Core Temp, and Orthos and have some fun.

Set up this way the DS3L dynamically adjusts Vcore and the mutiplier to save energy.

On my E6300 and the friends E2160 I have seen Vcore as low as 1.12V at idle, and around 1.28V at full load.

I can see this varying in CPU-Z. Use Orthos to add load and stop to remove loading on the CPU.

I can also watch the mutliplier vary. The lowest mutiplier used for a C2D is 6, so my E6300 varies between 6 and 7. The E2160 has a multiplier of 9, so it varies between 6 and 9.
Both are overclocked yet run relatively cool this way. The E6300 is running at 7 X 400Mhz and the E2160 is at 9 X 333.

That E2160 really rocks: 2997Mhz and 22C idle, all with he stock Intel cooler.

 

[rbk123]

My Q6600 does 3.2G on stock volts. Technically it's undervolted, actually, due to the V-droop of my board.