My QX9650

[CoolZone]

I have recently bought a QX9650.I want some details from you guys on what to expect from it since I am buying the DFI X48 T2R mobo in May.

The FBO/Batch is : L740A489T and it a sealed Retail box,packed in March 2008.


Thanks in advance.

[fallwind]

If we're talking 24/7 settings, my bet would be 4.0-4.1 on high end air. 4.2-4.4 with water. Mine does 4.2 on water with 1.44v

[CoolZone]

you have a batch similar to mine?

[BenchZowner]

That depends on what you mean.
If you're going for sub-zero cooling, like Dry Ice & Liquid Nitrogen, I'd say lay low... as most of those L740s max ( or should I say crap out ) at 5.3x GHz.
On air/watercooling most them ( L740s ) are decent clockers.

Let's say 4.3GHz with a ThermalRight Ultra-120 Extreme is a reachable target.

[CoolZone]

I'm going with the Thermalright Ultra 120 Extreme

[Shalamay]

What can one expect on phase?

[BenchZowner]

That also depends on various things.
Your luck, your phase change's performance ( cooling capacity & temps reached ) and what do you mean... 24/7 stability ? benching ?

[CoolZone]

can the people that post here tell us please their Batch code too?Thanks in advance.
Has anyone bought their CPU from the person that sells them on ebay for 800$ buy it now ! price?

[Leeghoofd]

I have bought a 740B159 from Ebay but I specifically requested the batch nr and a picture of the CPU (to verify the wafer nr too) till now a good CPU3.6ghz at 1.15Volts and 4ghz at 1.26V daily... way better results then my retail one lol

[thebanik]

can the people that post here tell us please their Batch code too?Thanks in advance.
Has anyone bought their CPU from the person that sells them on ebay for 800$ buy it now ! price?

I have, got to 5.41Ghz with 1.9V under LN2, maybe there is still some gas left in it if i get more courage to push higher volts. And this is without asking for any specific batch or checking the wafer no. Just plain lucky maybe. Did 4.6Ghz on TRUE as well.

[zoob]

Ends with a "T"!

It seems those ones are special..

[CoolZone]

mine ends with a T too,haven't tested it yet...hope it is good

[antari]

I have a L740A931T, packdate in late December '07. 4GHz @ 1.25V, have not gone higher as my cooling is somewhat poor.

[CoolZone]

any more users that can share their batch?What does that T really mean?

[TouGe]

I've got a L744B163 01/23/08 retail and its pretty mediocre compared to some around here. I can get 3.6GHz (9X400) @ ~1.26V (still testing for lower Vcore) but need 1.34V to reach 3.8GHz (9.5x400). Voltages are what's displayed in the BIOS/Everest, not what I actually set.

3.6GHz @ 1.26V <<<< 3.6GHz @ 1.40V which is what my x6800 needed.

[Leeghoofd]

Got the chance to quickly test a QX9770 L 803B248 ( C1 stepping) yesterday... was 3.6GHz stable at 1.2volts, 4ghz needed 1.32volts ot be OCCT 2.0 stable... heat output wasn't an issue with the Wc setup though was 5-8&#176;C more than my Qx9650... I really wonder where the extra cash goes into, are the new Qx97's beter resistant to abuse or what ? FSB was also around 470 maxed out... so no real improvement there on my board...

[CoolZone]

any speculations like at what voltage I could reach 4ghz stable?

[Leeghoofd]

Prolly in between 1.25-1.4 volts, if it can do eg 3.6 at 1.15volts then you are well on ya way to stay below 1.3volts for 4ghz...

Talking voltages under load here, not bios voltages, m P5K Dlx needs to be set at 1.2volts to do 1.15 under load (with laodlien calib enabled)

[cadaveca]

IT really varies. Good chips will take 1.3v or less, but I've seen 1.4-1.5v for 4ghz on a couple. MY L739"T" takes 1.272 for 4ghz.

[road-runner]

Mine is a L739B880 I have to set it on 1.45 in the bios to get 3.9Ghz prime stable...

[CoolZone]

has anyone found out what is with these 'T' chips?Are they more special or...?

[Leeghoofd]

General word for T CPU's is that they seem to clock pretty good ( 4ghz around 1.25Vcore level), maybe T stands for Turbo lol

[TouGe]

Prolly in between 1.25-1.4 volts, if it can do eg 3.6 at 1.15volts then you are well on ya way to stay below 1.3volts for 4ghz...

Talking voltages under load here, not bios voltages, m P5K Dlx needs to be set at 1.2volts to do 1.15 under load (with laodlien calib enabled)

Do you really recommend using LLC on these chips though? I thought that was a no-no w/ the 45nm parts. Is there a safe voltage level where LLC wouldn't cause damage to your chip from excessive spikes? I'm still not clear on the max safe voltages for these 45nm chips.

[Leeghoofd]

I have tested this P5K DLx for over hours now with different settings Touge, and have yet to see a spike... with the QX9770 I saw some ripples in the Vcore though nothing major... Qx9650 seems pretty flat...

Both CPU's tested at 3.6Ghz for a quick comparison (QX9770 test was interrupted after 40 mins):





At first the QX9770 was crashing with anything lower than 1.22volts at 3.6ghz, though after a night of folding I completed a few OCCT ( dunno why but it were OCCT1.0 runs so no voltage readouts) tests at only 1.2volts, do these CPU's need a burn in again ?

I know Anandtech doesn't recommend LLC but I think if you got a good mobo and good PSU then you will be fine... apparently they have found out that it seems to be CPU VTT that kills these CPU's if set too high (mine is always set at 1.3)

[TouGe]

That's good to hear. When I was testing with LLC enabled I didn't see any ripples at all but I thought that maybe the sample rate that OCCT uses might not be fast enough to catch them. When testing w/ no LLC I did see some ripples reported by OCCT

@Coolzone
I have no idea about this mythical "T". I'm just mad that I didn't get one.