Search the archive:
YaBB - Yet another Bulletin Board
 
   
 
Page Index Toggle Pages: 1
Send Topic Print
Help running DDR3 1600 on i7 965...SOLVED! (Read 647 times)
Dec 31st, 2008 at 5:41pm

raptorx   Offline
Colonel
There's too much confusion...
San Diego, CA

Gender: male
Posts: 434
*****
 
Just got the i965 and P6T Deluxe.  Great so far, but with my Corsair Dominators I can't run memtest v2.11 without getting errors at DDR3 1600.  

They run fine at they're SDP of DDR3 1333, 9-9-9-24.  DRAM bus is 1.64v and I know that's max.  There is an XMP profile but it puts the QPI voltage at 1.36 which is too high (in the purple).  I tried running at vQPI 1.25 and 1.268 volts and still errors.  Do I need more QPI volts?  

Is the XMP profile the one to use, is it safe?  Or should I use manual settings?  I'm just getting my feet wet here.

I haven't started overclocking the cpu yet-all those settings are on auto (3.2GHz @ 24x133).  I want to break in the cpu for a week or two before I start clocking.

Thanks for the help,
Jim
« Last Edit: Jan 1st, 2009 at 4:29pm by raptorx »  

Rampage II Gene, i7 965 4GHz
Mushkin Redline DDR3 1600
XP x64 SP2
ASUS Matrix GTX285
IP Logged
 
Reply #1 - Dec 31st, 2008 at 5:47pm

NickN   Offline
Colonel
FSX runs fine... the problem
is you or your system

Posts: 6317
*****
 
XP x64 SP2, 3 x 2Gb Corsair Dominators @ DDR3 1333, 9-9-9-24


IS that what they are rated for?

Did you buy the 1333 sticks or the 1600 sticks?

If you bought the 1333 product, you bought 1333.. there is no clocking with those and its why I posted the use of the OCZ or Mushkin 1600 product

DO NOT RUN VDIMM HIGHER THAN 1.65v!

If you bought the 1600 product and it wont run 1600 @ 1.65v on the specified timing then you need to see the Corsair people in their forums. It sounds like defects




1.90v vdimm / 1.1 vtt = dead after 10 hours (Engineering Sample CPU rev B)

1.65v vdimm / 1.20 vtt = ok forever according to intel
1.80v vdimm / 1.30 vtt = ok for 4 weeks
1.95v vdimm / 1.40 vtt = ok for 2 weeks
2.05v vdimm / 1.45 vtt = ok for 2 weeks
2.30v vdimm / 1.35 vtt = ok for 1 week (cold)


OK now DRAM VOLTAGE and QPI/DRAM VOLTAGE are 2 differnt things. The 1.35v is fine in the profile

keep the QPI voltage within 0.5V difference of the DRAM voltage and there’s no problem.

EDIT: THATS .5  not .05!
 
IP Logged
 
Reply #2 - Dec 31st, 2008 at 5:54pm

raptorx   Offline
Colonel
There's too much confusion...
San Diego, CA

Gender: male
Posts: 434
*****
 
I bought the 1600s.  Is it that I need to be closer to 1.35v vQPI?

 

Rampage II Gene, i7 965 4GHz
Mushkin Redline DDR3 1600
XP x64 SP2
ASUS Matrix GTX285
IP Logged
 
Reply #3 - Dec 31st, 2008 at 5:59pm

NickN   Offline
Colonel
FSX runs fine... the problem
is you or your system

Posts: 6317
*****
 


Sorry was editing my post... re-read
 
IP Logged
 
Reply #4 - Dec 31st, 2008 at 6:10pm

raptorx   Offline
Colonel
There's too much confusion...
San Diego, CA

Gender: male
Posts: 434
*****
 
Thanks Nick.  I'll run the XMP profile with 1.35 vQPI and test and report back.

I'm loving this so far! Smiley  Can't wait to really start clocking.  I've been reading a lot and looking stuff up so in a couple of weeks or so I should be safe to not fry anything.

-Jim
 

Rampage II Gene, i7 965 4GHz
Mushkin Redline DDR3 1600
XP x64 SP2
ASUS Matrix GTX285
IP Logged
 
Reply #5 - Dec 31st, 2008 at 6:12pm

NickN   Offline
Colonel
FSX runs fine... the problem
is you or your system

Posts: 6317
*****
 





Kill TURBO MODE and ALSO the CPU TM Function in the BIOS before setting up the clock

kill the CPU advanced TM crap and turbo mode and it should be stable in high clocks
 
IP Logged
 
Reply #6 - Jan 1st, 2009 at 4:29pm

raptorx   Offline
Colonel
There's too much confusion...
San Diego, CA

Gender: male
Posts: 434
*****
 
I ran the XMP profile and set the vQPI to 1.34v and the vDIMM to 1.64v.  Passed memtest.  I also disabled all the other cpu fluff and the spread spectrums.  I failed earlier because I had the vQPI too low. 

So I know the memory's stable @ 1600.

Now to the clock.  All I have to do is raise the multiplier and vcore, right (keep under 1.45v)?  No messing with the QPI or other stuff?  Run stable in stress testing under 78 deg C?

I want to run the 25 multiplier for a week or so (3.33GHz) before I go for 4GHz (after installing the TRUE).  I plan to not overclock the QPI unless you advise so.  I'll run 30 x 133 if I can.

Happy New Year! Smiley

-Jim
 

Rampage II Gene, i7 965 4GHz
Mushkin Redline DDR3 1600
XP x64 SP2
ASUS Matrix GTX285
IP Logged
 
Reply #7 - Jan 2nd, 2009 at 4:36pm

raptorx   Offline
Colonel
There's too much confusion...
San Diego, CA

Gender: male
Posts: 434
*****
 
NickN wrote on Dec 31st, 2008 at 6:12pm:
Kill TURBO MODE and ALSO the CPU TM Function in the BIOS before setting up the clock

kill the CPU advanced TM crap and turbo mode and it should be stable in high clocks


I should leave the cpu power saving features on though, right?  I read in the article you referenced in another post that the C1E and speed stepping work much better with the i7 then they did with the Penryns.  Is that correct?

-Jim
 

Rampage II Gene, i7 965 4GHz
Mushkin Redline DDR3 1600
XP x64 SP2
ASUS Matrix GTX285
IP Logged
 
Reply #8 - Jan 2nd, 2009 at 4:41pm

NickN   Offline
Colonel
FSX runs fine... the problem
is you or your system

Posts: 6317
*****
 


I turn it off but if you are stable with it running then leave it on

Its not going to change perf.. as long as you are not having any stability issues and no strange stutters associated with it.. leave it on
 
IP Logged
 
Reply #9 - Jan 2nd, 2009 at 5:16pm

raptorx   Offline
Colonel
There's too much confusion...
San Diego, CA

Gender: male
Posts: 434
*****
 
I have both the C1E enabled and the C-State enabled.  Is there a difference between the two?  Do they both go together?

Last Q in this thread. Smiley
 

Rampage II Gene, i7 965 4GHz
Mushkin Redline DDR3 1600
XP x64 SP2
ASUS Matrix GTX285
IP Logged
 
Page Index Toggle Pages: 1
Send Topic Print