Did you try reseating the heatsink? More than likely thats the problem. Or you don't have a good thin coverage of thermal paste.
Printable View
Thanks Ed! I am pretty sure the default on the P6T is 1.8 but I will find out for sure when I get home. :)
Chicken Patty, I sent ya a PM, check it out.
Nice clocks mate, my previous chip was from the same batch as yours, but it was faulty!!
Temps were stupid high at stock, and it could only do 3.8GHz, I spend over a week trying for 4GHz but it wouldn't happen..
Thankfully the e-tailer agreed to give me a replacement, now I have a nice chip!
Hmmm my 920 CPU doesn't like the 20x multi. at all, I couldn't get it prime95 stable @ 20x200 at all but now I finished 7 hours of prime95 with 21x195 with voltage of 1.28, I might increase it to 21x199 and see if I can lower the voltage.
Very nice clock and temps there Ejic! :clap:
http://i552.photobucket.com/albums/jj339/oggiet/oc.jpg
Know its not great speed but when had 200X20 had to run it 1.32V ...... with x19 its much easier to oc
Does the x19 (211x19) requires less volt than x21 (191x21)?
Damn nice voltages mate, are you using CPU/IOH clock skew, if so what values do you have them set to?
I have my CPU clock skew set to 100ps for my 4GHz o/c, it has helped me get both my Vcore and Vtt down quite a bit.
I am now stable with 1.2645v Vcore and 1.215v Vtt, whereas before I set clock skew to 100ps, I needed Vtt at 1.275v and 1.2875v Vcore, it has made quite a difference to my load temps having these voltages down as low as this, though I wish I could use low voltages like some of the guys on here.
No mate, that wasn't a typo, I'm running 1.300v CPU PLL, and my boards default is 1.800v, I read on EVGA forums that lowering this setting can actually help with stability, so I started playing around with it to get my 4.2GHz o/c stable, and I got it right down to 1.300v, when it was at 1.800v my overclock was unstable, now it is perfectly stable.Quote:
Originally Posted by ed.howell
I know several other guys using 1.300v CPU PLL as well.
Have you tried lowering it down even further than 1.82v?
Hi mate, I had quite a hard time getting my 20x200 o/c stable as well, whereas I could get 21x191 stable without any trouble at all.
But I kept on persevering with it, and I finally got it stable.
I have found that using high voltages can also cause instability to, when I first started trying to get 20x200 stable, I was having problems, so I tested with Vcore at 1.35v and Vtt at 1.375v, and it just wouldn't pass Prime95 blend test or LinX at all, it kept giving me BSOD's, but it is now perfectly stable with only 1.275v Vcore and 1.215v Vtt!
It's is all about finding your chips sweet spots, I'm not 100% sure, but I think that some chips like to have specific distances between Vcore and Vtt, some like wider distances, while others prefer them to be closer together.
I've had 4 of these chips, and my first 2 seemed to like the distance between Vcore and Vtt to be closer than the 2 chips I have here atm.
@oggie, have you just got your CPU in your loop, or do you also have your GPU(s)?
CPU is in one loop, gpu and NB in other ;)
Hi mate,
My apologie for the delayed response....
Nope 4.5GHz HT on stable is the Max capability on Air for linx/priming when its on my table with push and pull config...still the Core temp reaches about 88-92C after an Hour or so :mad:
I can also boot in to windows with 4.6GHz and more but temp not suitable for benching.... looks like with water 4.7GHz could become my 24/7 settings... though I dont see any practiacal use of that speed in current applications.
24/7 settings on Air is 4200MHz HT on inside the case reasonable temp after benching... see the pix that will tell you all my settings... :up:
http://img200.imageshack.us/img200/505/4200mhz0.png
http://img193.imageshack.us/img193/4205/4200mhz1.png
imo,
Why are you running Linx with only 1624MB if you have 6GB RAM? Click on the "all" button and check your stability again please, I doubt it will be stable with only this vcore.
Just redid my voltages a bit :)
http://i31.tinypic.com/2d139kp.jpg
4.2 HT on
EDIT: here it is
http://i30.tinypic.com/9uqlxt.jpg
You guys are lucky with that voltages. Is LinX as good as Prime95 or OCCT for stability testing? My first attempt at 4.2Ghz @1.4V and I can only back down to 1.39 in order to be OCCT stable.
Btw, imo, I think your temps are a bit high @ that voltage since this is mine @1.4V
http://img198.imageshack.us/img198/2...pboard011j.jpg
It was Auto to Prime95
CPU/IOH clock skew changes by Chip
I think the combination to be 100/0 ・ 200/100 ・ 300/200 ・400/300 ・ 500/400
QPI/DRAM Core Voltage
CPU PLL Voltage : 1.80V
CPU Diffrential Amplitude :
IOH Voltage :
Load-Line Calibration : Enabled
CPU Clock Skew :
IOH Clock Skew :
It is necessary to verify it in detail.
It might be useless only to increase Vcore
The combinations of each voltage of each CPU
To save time as I wasnt showing my stability @ 4200MHz with 1.23v since I have tested that before and know its stable...I was more like showing the temp I get and the settinggs I used in response to his question. :rolleyes:
Any ways heres a glimpse of all memory used... nearly about 1 hour test all available memory used... that I have dedicated to your doubt...:ROTF:
http://img19.imageshack.us/img19/1774/4200k.png
Well.. what can I say. That's a really impressive chip, congrats :P
@blunt, what did you do to get 20x200 stable?
@imo, what is your CPU batch?