Nehalem Overclocking @ XS

[Charles Wirth]

Update

Voltages

vcore ~1.5v, this is a "leaky" high IDV chip, scaling up with higher voltage is possible at a cost of runaway amp phenomena.
vmem ~2.2v for benching 2v on Vapo, compensate voltage with cooling to be safe.
vqpi ~1.35v depending on the board/bclock, upwards of 1.55v if I am pushing bclock, I run a low bclock so I usually leave it on stock.
vmem is the secret to uclk, 2.2v gets you 4.7Ghz ulck. uclk is also known as uncore.

general ratios

Stock QPI on i7 965 is x48 @ 133 bclock, if you raise the bclock to 160ish you would need to drop qpi ratio to X44 to continue raising the bclock to 180ish, drop qpi again to X36 again and that will get you out the theoretic max bclock of 220. That is not set in stone, some chips/board combos have exceeded these numbers but this rule is pretty close to what you will see on C stepping.

The uclk ratio I prefer to use is (the lowest possible) 1:2 of my ram speed (uclk runs at twice the speed of memory) and at lower speeds there are plenty more ratios to scale higher without running the sick memory speeds that I do, this is a huge gain in performance but for me it requires a lot of vmem and high end memory to run upwards of 2300Mhz (4700Mhz uclk) the memory controller has exceptional is its stability at this speed, I have not reached top speed and it will take even faster ram and more vmem to achieve that.

Runaway amp phenomena, I hook an amp meter to the four yellow wires on eight pin power cable that connects near the CPU. Watching this you will see that yes you can keep scaling up with more voltage but your chip has a limit with all cores loaded that it will runaway on amps till it crashes. I know exactly what range to stay in with cores loaded to know not to hit this runaway threshold. Keeping vcore down and disabling HT lets me run at the chips max speed, HT has an added power load on the CPU, disabling it helps gain critical Mhz.

[dinos22]

good info charles

let us know when Intel figures out what causes runaway amps
this is something that needs to be metered it seems

will have to get a meter
if you could link to the one you use it would be handy for members here

good insight into OCing in general
i'll make sure to go over it when i finally start my testing soon

[T_M]

without having measured it i know exactly what you mean about vcore being too much
im still working with my better chip to find the right volts to get highest clock without crashing when i high load hits.

uncore is really handy for PI efficiency

[[XC] gomeler]

Note sure if you can say its improvement, but here's my result from last night with only an Intel SSD (no iRams etc):

http://www.hwbot.org/result.do?resultId=792285

Would be helpful if I had a compare URL so I could see where exactly Core i7 is showing improvements. I suspect though the only improvement you will see will be in the last multi-threaded test with a large storage array. The memory access at 19.x MA/s is impressive and the extra available CPU threads for the actual HDD accessing will help the virus scan score. PCM05 only scales to 4 threads I believe so those 4 HT cores will sit idle or help with background stuff like accessing the HDD array.

[Sampsa]

Runaway amp phenomena, I hook an amp meter to the four yellow wires on eight pin power cable that connects near the CPU. Watching this you will see that yes you can keep scaling up with more voltage but your chip has a limit with all cores loaded that it will runaway on amps till it crashes. I know exactly what range to stay in with cores loaded to know not to hit this runaway threshold. Keeping vcore down and disabling HT lets me run at the chips max speed, HT has an added power load on the CPU, disabling it helps gain critical Mhz.

Hmm, what method are you using to measure the current? Cutting the wires and adding multimeter in series or clip-on ammeter like this http://www.hotektech.com/GMCMetraclip70.htm from one of the yellow wires or from all the yellow wires? What kind of values you see? And what value is the limit (runaway threshold) when cores are loaded? Is it some Intel spec or experimentally figured value? Thanks

edit: looks like you can use LTS 25-NP current sensor:

http://www.silentpcreview.com/article313-page4.html

Product page: http://www.lemusa.com/hq/en/componen...0.37.19.000.0/

[slash777]

My i7 920, VID : 0.975

[dinos22]

Hmm, what method are you using to measure the current? Cutting the wires and adding multimeter in series or clip-on ammeter like this http://www.hotektech.com/GMCMetraclip70.htm from one of the yellow wires or from all the yellow wires? What kind of values you see? And what value is the limit (runaway threshold) when cores are loaded? Is it some Intel spec or experimentally figured value? Thanks

edit: looks like you can use LTS 25-NP current sensor:

http://www.silentpcreview.com/article313-page4.html

Product page: http://www.lemusa.com/hq/en/componen...0.37.19.000.0/

he doesnt do any modding/cutting

just a cheap clip on meter

[[XC] gomeler]

Are we positive the CPU pulls all it's power only from the 12v line? No power is drawn from the 24 pin ATX connector or is it "negligible"?

edit: we know the CPU pulls power from the 24-pin connector, these boards will run without the 12v line plugged in

My i7 920, VID : 0.975

Something special there? My E8600 will run with similar vcore and it's been abused to no end

[T_Flight]

Thanks Fugger for the post on "Runaway Amp Phenomena". I learned something new today.

[makaka]

huh ............ 24 k of memory bandwidth are u kidding ???????.
Congrats on those results , i hope that core i7 will be available here soon , i need to do some benchies , i am tired from LGA 775:-)

[Charles Wirth]

Here is a pic of the amp meter attached to the 12v lines.

This is an example of where to attach, do not purchase model shown.

[eva2000]

Fugger was runaway amps only on Gigabyte board or other brands ? could it be related to Gigabyte X58's voltage controller/load line calibration not working and needing Gigabyte to revise the board ??

[freshy]

Fugger was runaway amps only on Gigabyte board or other brands ? could it be related to Gigabyte X58's voltage controller/load line calibration not working and needing Gigabyte to revise the board ??

i hope the ud5 is a good board ? i just bought one

[pinto]

New toy, watercooled on P6T deluxe:

[Charles Wirth]

All boards, first seen on Smackover day one.

More feedback from people, we will probably see the F1 pot being able to run into higher amps than cascade.

It would be good to know what others are peaking at.

[T_M]

Unfortunately i dont have one of those puppies, so anyone else?

[Sampsa]

Unfortunately i dont have one of those puppies, so anyone else?

I'll go check local stores today if I can find clamp-on current meter, should get new 965 chip to play with tomorrow and LN2 on friday

[Sampsa]

My local shop had this for 100 euros:



http://www.multimeterwarehouse.com/MS2108.htm

They need to order it so it will take couple of days. Should do the job right?

[blossa]

It should be ok. 0.1A intervall should be good enough atleast to see how the trend is.

Nice function that "clamp on frequency function". I have never seen that before. Another good function is capacitance measuring. However, i am not sure if it has that as it is not mentioned in the detailed specification... Peak function might get usefull aswell.

[LardArse]

My local shop had this for 100 euros:



http://www.multimeterwarehouse.com/MS2108.htm

They need to order it so it will take couple of days. Should do the job right?

woahhh I think I can get one for 10 euros with lesser bling

[Slovnaft]

ahaha, the other day I accidentally bought an hvac clamp meter, now i find out it'll come in handy!

might this 'runaway amp' phenomenon be related to the new power shifting algorithims?

[Praz]

Runaway amp phenomena, I hook an amp meter to the four yellow wires on eight pin power cable that connects near the CPU. Watching this you will see that yes you can keep scaling up with more voltage but your chip has a limit with all cores loaded that it will runaway on amps till it crashes.

What average amperage are you seeing before runaway happens?

[TiN_]

Aren't that clamp meters only AC? AFAIK they able to measure only AC currents, not DC +12V.

[Charles Wirth]

"AC/DC Clamp Meter"

[Sampsa]

Aren't that clamp meters only AC? AFAIK they able to measure only AC currents, not DC +12V.

There are models which support only AC, but also models supporting both AC/DC.