Lapping QX9650 Project (and max OC on air)

I've seen a couple of threads about people wanting to lap QX9650, but didn't see any pics, so decided to lap mine (along with TRUE) and post some pix for everyone to enjoy. Plus would like to share my max OC of the processor on air.




Let's roll!

http://i161.photobucket.com/albums/t...9650/TRUE1.jpg
Getting ready to lap. Grit from left to right: 400, 800, 1000, 2000, 2500

http://i161.photobucket.com/albums/t...9650/TRUE2.jpg
400

http://i161.photobucket.com/albums/t...9650/TRUE3.jpg
800

http://i161.photobucket.com/albums/t...9650/TRUE4.jpg
1000

http://i161.photobucket.com/albums/t...9650/TRUE5.jpg
2000

http://i161.photobucket.com/albums/t...9650/TRUE6.jpg
2500


And now the QX9650 (L739A731)
http://i161.photobucket.com/albums/t...x9650_lap1.jpg

http://i161.photobucket.com/albums/t...x9650_lap2.jpg
400

http://i161.photobucket.com/albums/t...x9650_lap4.jpg
800

http://i161.photobucket.com/albums/t...x9650_lap5.jpg
1000

http://i161.photobucket.com/albums/t...x9650_lap6.jpg
2000

http://i161.photobucket.com/albums/t...x9650_lap7.jpg
2500

http://i161.photobucket.com/albums/t...x9650_lap8.jpg
The job well done

http://i161.photobucket.com/albums/t...x9650_lap9.jpg
Mirror finish 1

http://i161.photobucket.com/albums/t...0_lap_refl.jpg
Mirror finish 2



Results:

In all cases, the setup was left for 12h idle and 3x 1024M tests in wPrime
*Note, the cpu is running at 4.2GHz

http://i161.photobucket.com/albums/t...ingresults.jpg
Didn't bother doing Shin Etsu since it proved to be not as good as AS5... (*typo)
Also to note about AS5 load -> Real temp 2.21 showed 53°, Asus Probe 2 showed 55°.
So I took the average of the two.



With the temps lowered a bit, I was able to beat my previous record of 4.55GHz and reached 4.7GHz on air!!! :eek:

http://i161.photobucket.com/albums/t...max_on_air.jpg




And of course, CPU-Z Verified.

http://i161.photobucket.com/albums/t...z_verified.jpg




Yep, so that was my little QX9650 project. :cool: :up:



_______________
________________
_________________

veery nice OC

Verry nice done !

lower that voltage or you will end up with a dead chip with a voided warranty

Personally I never lap any chip I cant afford to replace without hurting my bank acct.

BTW Remount with no TIM, on a good lap job you usually get better temps with just metal on metal :)

Or you can use that liquid metal stuff :D

It takes a brave man to lap a $1k+ chip. Nice work tho !

Nice lap job. It takes guts to lap a QX9650. Good temps and great OC.

nice lapping job you did there with a QX9650.

Good results ,
Now you need to try this CPU on DICE/LN2 .
You'll get stunning results with this cooling

Away for some sanding paper :)

Great results!

Awsome results!! It would also be really nice to see how much of a difference MX-2 would make as well. You might get another 1-2c improvement.

Quote:

---

Originally Posted by Cupcake

lower that voltage or you will end up with a dead chip with a voided warranty

Personally I never lap any chip I cant afford to replace without hurting my bank acct.

BTW Remount with no TIM, on a good lap job you usually get better temps with just metal on metal :)

Or you can use that liquid metal stuff :D

---

Is it really a good idea to not use any TIM at all unless the lapping is done professionally (that is, by machine)? Seems like human error in lapping necessitates using at least a small amount thermal paste.

And MartinM, speaking of liquid metal, says he laps his IHS whenever he wants to remove the stuff. Is it really that hard to remove?

Very good! Ive seen sumone do it 1.6V 5ghz on air! Can you try that out? Im sure you can do it. Youre pretty good! Dang. What an idol. 5ghz please ^^.. and show the world what a Striker II Formula bed can do!

And some orthos/stressprime screenies..

Nice results. Does lapping really helps do get lower temps?

good result! nice lap work MD and great OC :up:

Quote:

---

Originally Posted by Cupcake

lower that voltage or you will end up with a dead chip with a voided warranty

Personally I never lap any chip I cant afford to replace without hurting my bank acct.

BTW Remount with no TIM, on a good lap job you usually get better temps with just metal on metal :)

Or you can use that liquid metal stuff :D

---

Bad idea in my experience, no matter how smooth and shiny you get everything there are still microscopic pores. I stand to be corrected though.

Thanks guys, there is nothing more awarding than hearing positive feedback on a project with many hours of dedication.

Quote:

---

Originally Posted by Cupcake

lower that voltage or you will end up with a dead chip with a voided warranty

---

Voltage is raised so much only to achieve the max OC. From my sig, you can see that I run 4.2GHz at 1.375V very stable. Oh, and warranty is already voided - lapped cpu?! :p:

Quote:

---

Originally Posted by Cupcake

BTW Remount with no TIM, on a good lap job you usually get better temps with just metal on metal

---

Yes, I will try that, as well, as Xphobe suggested, with MX2, and will post the results.

Quote:

---

Originally Posted by lior307

Now you need to try this CPU on DICE/LN2. You'll get stunning results with this cooling

---

I'm planning on moving into phase change and building a custom system this summer. I'll definitely keep ya'll posted. :yepp:

Quote:

---

Originally Posted by Thakkerson

Very good! Ive seen sumone do it 1.6V 5ghz on air! Can you try that out? Im sure you can do it. Youre pretty good! Dang. What an idol. 5ghz please ^^.. and show the world what a Striker II Formula bed can do!

---

Where did you see that? The very max I was able to squeeze out of my proc, was 4.75GHz, but it was so unstable that I didn't even have time to take a screeny. You see, as I found out, there is no way to score a higher OC at 1.6V on air, because even at my 4.7GHz, idle was 50° and load is 72°. If I raise either the volts or the freq, temps are going insane and windows crashes.




P.S.


I forgot to mention that I put NB and SB on AS5 as well as my 8800GTX.
From the screen with the overclock you see the Everest temps in the corner, and (from left to right) first is 49° diode and 40° gpu. This made it possible for a 666/2100 overclock.


http://i161.photobucket.com/albums/t...0/8800gtx1.jpg
http://i161.photobucket.com/albums/t...0/8800gtx2.jpg

Quote:

---

Originally Posted by Thakkerson

Very good! Ive seen sumone do it 1.6V 5ghz on air! Can you try that out? Im sure you can do it. Youre pretty good! Dang. What an idol. 5ghz please ^^.. and show the world what a Striker II Formula bed can do!

---

No way ... :shakes: On air ? Raising the Vcore after a certain point will be useless when aiming at high overclocks on air ... It does no more, than " killing " the chip ...

There are some brave guys though (like Benchzowner) who feed their wolfies with 1.8 Volts on air ... :yepp:

@_MD_ : Nice job man :up: I've heard that lapping the cpu, will only be beneficial on air or water cooling ... It's pointless when using subzero cooling, and sometimes it can even make things worse ... ;)

awesome job MD. lapping and oc. :up: this belongs here, you know what i mean.

very nice man!! good too see someone who lives nearby pushing it to the limit. not to many people in our area going xTReMe it seems.

Good going fella. :up:

Quote:

---

Originally Posted by _MD_

Where did you see that? The very max I was able to squeeze out of my proc, was 4.75GHz, but it was so unstable that I didn't even have time to take a screeny. You see, as I found out, there is no way to score a higher OC at 1.6V on air, because even at my 4.7GHz, idle was 50° and load is 72°. If I raise either the volts or the freq, temps are going insane and windows crashes.

---

I gather that was 1.62V BIOS / 1.54x Windows... then you're DTS temps are being reported minimum 10-15C lower than they really are at load. 1.5V on a QC 45nm chip, don't expect below 50°C idle at 20°C socket ambients, with a TRUE, on air and far above 72°C load. It won't be possible (idle is in one case, if you have CE1 enabled). It physically can't when 51°C is full load at 23C socket ambients with a TRUE using a 152CFM Delta and a 1.2V/3G 65W TDP 45nm DC chip, which is half of this chip TDP at just stock. I'd be careful how much you roast it. :)

Nice work dude.

Very nice work :up: - any rig pictures?

NOw thats sweet! I've never seen a polish so pretty.... Good work!

Wish I could understand the way sand paper is graded when it gets past the ultra fine stage.

I've always used this kit -
http://www.kustompcs.co.uk/acatalog/info_3719.html

But I've just taken a scalpel and removed the IHS from the CPU. I've never lapped the bare core - I don't know if it's possible without damage.

But I'm going quad, so I'll be getting a 940 BE, which will need to be lapped. So my question is how did you hold the CPU to lap it?

Quote:

---

Originally Posted by Cupcake

lower that voltage or you will end up with a dead chip with a voided warranty

---

I think the lapping already covered the voided warranty :)