SimC33

*UPDATED*


1. Flyingduck, your post should be good now on the list...

2. ndtran and kiki, I really appreciate the post and participation, but could you do it again with a problem size of 25k+ then post again so I can add yall?

3. chememan, congrats on your new system being 4.0 efficient...I notice it was with HT off. I suggest and encourage you to have some fun and get the PC to 4.0 with HT ON...give it a shot. I bet you can make it work.

4. Can everyone PLEASE remember to post your results and INFO like the example shows in the original post please. Im still digging around fixing everyone's initial info they give me.

Thank ya!!!

SimC

Thank you SimC for all of the hard work and time you put in to this!!!
A real class act!!

chememan

SimC33 

3. chememan, congrats on your new system being 4.0 efficient...I notice it was with HT off. I suggest and encourage you to have some fun and get the PC to 4.0 with HT ON...give it a shot. I bet you can make it work.

I have been successful with HT on but I intentionally did the screenshot with HT off because it seems to provide the best performance with this benchmark.  To me, it is frustrating to put up a screenshot of the system showing 4.2ghz with dismal Gflops that are lower than folks with lower speeds and HT off.  I am personally looking for the maximum Gflops.  I would rather see 70+ Gflops than 5.0.  If HT gave a benefit beyond more heat/stress, I would use it.  At this point, I have not done anything where HT gives a benefit.
 
 

Fair enough and well put...if you ever do choose to decide to run the test with HT on and post a screenie, ill gladly add you to that list as well

Badass ndtran! Im glad you took the time to give it a shot and didn;t complain...im adding you now

thanks...nice to see I am one of the few with c0 stepping.  Anyhow I want to try higher but seem my corsair dominator 12800 doesnt like anything above where its at now.  I see many review says the evga x58 doesnt like this memory.  Any input?  It doesnt even like being at its rated 1600mhz at 8 8 8 24. 

I wouldn't mind setting the 965/C0 highest clock in the list(by ~30mhz) but I'm not cool enough because I have an ASUS board

Updated to the list kiki! Congrats

Updated data:
 
Chememan 
Xeon W3540 batch 3851A   
Cpu Clock 4313.4mhz
Voltage: 1.282 w/o vdroop
Idle temps: 40/37/36/33
Load temps: 66/62/62/59
HT off , water cooled
I loosened up the memory some to eliminate the possibility that it would interfere with the OC and it seems to have hurt the Gflops but the overall time was faster.  Does this make any sense to anyone?  Upped the CPU by 100+ mhz and lost one Gflop but finished 3 minutes faster?
 
I am not meeting any resistance and intend to go to 4.4 on the next attempt and then will see how far I can trim back the voltages and memory settings without killing the stability.  For now I plan to leave things as they are until I reach my final clock.  This is time consuming and I wish I had more time to play with this.
 
Doubt that I will try much beyond 4.4 for 24/7 since I see little benefit.  What say the experienced folks?
 

 

I noticed that memory speed plays a large part in the Gflops measurement that LinX calculates. On my Q9650 machine, I changed my memory speed from 1600 to 1800 and gained about 20% on the Gflops.  That was a ~20% Gflop increase on a ~12% increase in memory speed.  On the Xeon machine I had been running the memory at sticker 9-9-9-24-35-1T and running the memory well below the rated speed.  Once I started to get close to the rated speed, I changed to the XMP profile of 9-10-10-25-32-2T  and may have even loosened that up some too, thus the loss of speed. 
 
Not really sure why the XMP profile is different from the sticker.
 
 

awalleyeguy

It seems that at some point performance stops increasing (may even decrease) as much as speed, but varies from chip to chip. Speculation of course.

Peligro
Core i7: 920 D0 Batch #:  3849B040
CPU clocked at:  4.2105ghz HT on/ Turbo on
Voltage: 1.376V
Idle temps:  37/35/37/33
Load temps: 81/80/79/75
 
Finally got some numbers I am happy with.  Gonna turn HT off and see what this guy can really do.
 

slyver08

Slyver08
Core i7: 920 D0 Batch #:  3939B100
CPU clocked at:  4.2ghz HT on/ Turbo on
Voltage: 1.411V
Idle temps:  45/41/43/41
Load temps: 97/97/91/89

Thermalright Venomous X /w 2 San Aces @ 1000RPM > 3000RPM (the highest temps were at 1000RPM, at 3000RPM temps went down to ~low 90's on all cores) I'm gonna need watercooling for 24/7 heavy usage...:D

Do you really need all that much vcore for 4.2? I reach 4.0 at 1.32vcore never mind 1.42...

shaolin95

I am finally in! :-)



Thanks

You forgot all the other info he requested in the OP.

Here's mine (finally) :D


ReVolut1on
Core i7 920 DO Batch#: 3850A807
CPU clocked at: 4189.7 Mhz
Voltage: 1.28750V W/O Vdroop
Idle temps: 36C
Load temps: 85C

lcabezas
Core i7: 920 D0 Batch #: N/A
CPU clocked at:  4.315ghz HT on/ Turbo on / Watercooled 
Voltage: 1.376V
Idle temps:  45/43/44/44 (I started Real Temp right after I started LinX so it doesn't really show my Idle temps:
Load temps: 89/86/85/84

I was able to get it stable with lower voltages but I didn't take a pic. This is the pic I took with higher voltages. I also passed it on 210blck but I didn't take a pic of it. At least I'm in right?

Core i7: 920 D0 Batch #: 3913B222
CPU clocked at:  4.110ghz HT Off/ Turbo On
Watercooled Bran New Corsair H50 
Voltage: 1.4000V (without Vdroop)
Idle temps:  34/29/31/27
Load temps: 69/65/65/61
 

Hey everyone, I'm flying into Daytona for the weekend to go watch the Rolex 24 thats going on. I will update the list come Monday upon my return to the civilized world. Until then, that is all. Have a great weekend everyone!!!


SimC