Official MSI P7N Diamond (780i) Discussion/Review/Overclock/Guide/BIOS Thread

[pacific vortex]

yes we have boot up with p07 ...everything works fine...raid, 8800gts 512 in sli with 174.74 drivers, 6 usb devices plugged in and all work at boot up, windows vista 64 with sp1...now to really see what this board can do will post soon

[loonym]

Nice to see people reporting half multis working with this bios. Just as Aaron said C1E must be disabled as well as EIST. In advanced bios features>cpu feature, press F4 and hidden menu will appear with this option. Also, I made sure pci-e bus frequency was at default before reporting the vid card oddness.

[NIVO]

ok thanks for the help that did it!!! Just a question here...why are those features hidden? Wouldnt it make sense to keep them shown instead of using F4? Undocumented things give folks headaches Now i can push on for higher goals.

[pacific vortex]

just passed first occt run at 470 fsb again without any major tweaking 1:1 ratio, nb to 1.525v ....now to try prime 95... cant get 475 to boot in windows yet would give me a 3.8gig o/c.... pretty sure the cpu can do it...if anybody wants to share with me the vista secret for taking a screenie ill gladly post pics...

[loonym]

Just a question here...why are those features hidden?

I think it's to do with MSI's large oem market.

[AaronYuri]

Not sure why MSI do it.

C1E Support is basically EIST but hardware-based. EIST is software-based (Windows power profiles etc). C1E will force Speedstep at all times.

Glad to see everything is working out for you all. Hopefully the same thing will come my direction.

[dogg]

After some problems with cmos after flashing bios .116 /thx to loonym for patience with me , this was nothing I could do by try and error way as usually /, now I'm very pleased by this version... No more problems with usb devices... My G15 keyboard and G9 mouse workin right after restart /so I can get to bios, which was not possible before, I needed to use PS2 old one/..Restart is not taking so long time like before /there was 5-10sec pauses between beeps and now beeping fast one by one so I think there are no more problems with mapping hardware/.... Looks like a Final Candidate /using apple talk/ for official release..

What I'm still not able is run system with CL4 memory settigs /which should be default with these RAMs / and overclocked cpu at the same time... Auto setting works, but result is 5-5-5-18 T2...Default is 1.8V and I've set 2.1...But probably there's problem with Geil sticks /they were suspicious cheap/....

So, some first results..
CPU 1.32V /CPU VID 1.0375V -readed frome Core temp application/
NB 1.35V
RAM 2.1V
rest of voltages on auto

Eist and C1 disabled /by the way, nevermind, that those cpu setting are hidden, because they are disabled by default, probably because Eist is enabled by default/

FSB 420 MHz
Muliplier 9x /never used lower, but I've had a look and there's possibility to use a half multipliers as well.../

So I'm on 3780 Mhz... /3.8G was my goal so I'm close and that was first try with this bios/
I'm using Linked setting so my RAM is 787MHz CL5 T2 now, what makes 15:16 ratio... manual setting to 800 cause just black screen...
As I mentioned, cannot change RAM setting without unstabble behaviour...
But with actual prices of DDR2 I see no problem to try some other...

Runned Everest stress test for about a hour without error and don't remember so nice flat line of cpu usage on 100% before...

3D Mark06 on default 1280x1024 no AA wit score 16879
1680x1050 4 AA /quality 4/ Anisotropic F /level 10/ with score 12402

I remember higher scores with 1.0 bios version but with better RAM settings on 4-4-4-15 T1 with 2.1V power setting, anyway for longer use too unstable... Really don't want see bsod 3 times a day..
I didn't try any game now / usually show more than tests..

Do somebody have some link with infos about FSB holes...? This is absolute unclear to me, what's this about and don't want to fill this thread with something, that's clear to others...

Anybody with E8400 on this board..? /edges of voltage settings for CPU and NB?/


---------------------------------------------
MSI P7N Diamond (780i)
E8400 Penryn on air (Zalman CNPS8700NT)
2x2G Geil Evo One DDR2 CL4 800MHz
2x MSI 8800GT 1GB OC in SLi
2x 150GB WD Raptor
750GB Seagate Barracuda
Power: Coolermaster M series 850W
Logitech G15 Keyboard
Logitech G9 Mouse

Vista 32Bit Ultimate

[x-built-Stan]

Dogg,

I had an E8400 installed for awhile. Check the board for my results. I was at 4.0GHz.

[Overspeed]

Hey everyone, I have new news. X-built, I would value your input on this especially. As previously reported, I have never been able to get a good OC out of my E8400 (see more specs in sig). I haven't given up on it thought. Tonight I tried some new things, but first some history:

Original BIOS - No OC ability
1B4 Beta allowed 3.4Ghz max
Recently updated to the .116 beta, still no OC past 3.4Ghz.

Some feedback I have been given is that having 4 dimms causes the NB to do more work, so tonight I took apart the rig and pulled dimms from slots 2 and 4, leaving me with just 2Gigs total. Rather ambitiously, I tried right away for 4.0Ghz using +0.2v vcore, 1.4v NB 2.10 dimms, and 25 FSB term. Bios would post but no windows load. I tried +0.25 on vcore same thing, so I guess the CPU volts wasn't an issue. So I decided to try something easier... the E8400 has a FSB hole at 1600 so I decided on 1666 and 833 ram. Set vcore to +0.2, Nb to 1.4, dimms to 2.20 and FSB term to 63. Guess what, windows loaded and thus i'm typing now.

So what does this tell us? I have NEVER been able to reach 3.8Ghz, not even close, that said I have not tested stability yet, but just the fact windows loaded is a big change. So was it the removal of 2 dimms? overcloking my remaining 2 dimms at 2.20v? the higher FSB term? some other factor? I really can't explain it. But at last now I have hope with the acheivment of 3.8Ghz. I would love some advice on where to go from here to reach the magic 4.0. I would also value some advice on how to test stability properly. I have SP2004 readdy to go but am unsure of options and settings.

Update: I decided to run some SP2004 tests. I ran 2 instances, one on core #0 and one on core #1. I set to blend testing. Both instances stopped at 5mins36secs with errors reported. My Vista Gadgets sidebar app also crashed, but windows remained operational. If it helps, here is the SP2004 log:

####################
More than one CPU detected.
You should run two copies of SP2004 simultaneously.
More than one CPU detected.
You should run two copies of SP2004 simultaneously.
Type: Blend - stress CPU and RAM Min: 8 Max: 4096 InPlace: No Mem: 1790 Time: 15
CPU: 3750MHz FSB: 333MHz [340MHz x 11.0 est.]
4/1/2008 11:59 PM Beginning a continuous self-test to check your computer.
Press Stop to end this test.
Test 1, 4000 Lucas-Lehmer iterations of M19922945 using 1024K FFT length.
Test 2, 4000 Lucas-Lehmer iterations of M19922943 using 1024K FFT length.
Test 2, 4000 Lucas-Lehmer iterations of M19922943 using 1024K FFT length.
Test 3, 4000 Lucas-Lehmer iterations of M19374367 using 1024K FFT length.
FATAL ERROR: Rounding was 0.46875, expected less than 0.4
Hardware failure detected, consult stress.txt file.
Torture Test ran 5 minutes 36 seconds - 1 errors, 0 warnings.
Execution halted.

Test 3, 4000 Lucas-Lehmer iterations of M19374367 using 1024K FFT length.
FATAL ERROR: Rounding was 0.5, expected less than 0.4
Hardware failure detected, consult stress.txt file.
Torture Test ran 6 minutes 30 seconds - 1 errors, 0 warnings.
Execution halted.
#####################

Cheers.

[AaronYuri]

Overspeed,

Try running the small FFTs test, as it will stress only the CPU.

[NIVO]

have the cpu at 3.2ghz stable and running ram 1000 at rated spec. Passing OCCT test of 1 hour, now for phase 2 of stability testing

[pacific vortex]

anybody mess around with the nb gtl yet???

[silent-circuit]

anybody mess around with the nb gtl yet???

It seemed to be the trick to gaining stability at any clock above ~3.3Ghz for me, but I never managed to dial it in.

[NIVO]

well cant seem to get stability dialed in just yet on ram at higher speed(fails at 1hr 15min OCCT test) so I have reverted back to 3.2ghz 800mhz ram settings. OCCT pass 4 hour stability test So this will be my base point to push further from.

[theYipster]

It seemed to be the trick to gaining stability at any clock above ~3.3Ghz for me, but I never managed to dial it in.

Some boards don't seem to need it, while others do. Loonym has had much success with the voltage. He may have some dial in ideas.

[Overspeed]

Hey everyone, I have new news. X-built, I would value your input on this especially. As previously reported, I have never been able to get a good OC out of my E8400 (see more specs in sig). I haven't given up on it thought. Tonight I tried some new things, but first some history:

Original BIOS - No OC ability
1B4 Beta allowed 3.4Ghz max
Recently updated to the .116 beta, still no OC past 3.4Ghz.

Some feedback I have been given is that having 4 dimms causes the NB to do more work, so tonight I took apart the rig and pulled dimms from slots 2 and 4, leaving me with just 2Gigs total. Rather ambitiously, I tried right away for 4.0Ghz using +0.2v vcore, 1.4v NB 2.10 dimms, and 25 FSB term. Bios would post but no windows load. I tried +0.25 on vcore same thing, so I guess the CPU volts wasn't an issue. So I decided to try something easier... the E8400 has a FSB hole at 1600 so I decided on 1666 and 833 ram. Set vcore to +0.2, Nb to 1.4, dimms to 2.20 and FSB term to 63. Guess what, windows loaded and thus i'm typing now.

So what does this tell us? I have NEVER been able to reach 3.8Ghz, not even close, that said I have not tested stability yet, but just the fact windows loaded is a big change. So was it the removal of 2 dimms? overcloking my remaining 2 dimms at 2.20v? the higher FSB term? some other factor? I really can't explain it. But at last now I have hope with the acheivment of 3.8Ghz. I would love some advice on where to go from here to reach the magic 4.0. I would also value some advice on how to test stability properly. I have SP2004 readdy to go but am unsure of options and settings.

Update: I decided to run some SP2004 tests. I ran 2 instances, one on core #0 and one on core #1. I set to blend testing. Both instances stopped at 5mins36secs with errors reported. My Vista Gadgets sidebar app also crashed, but windows remained operational. If it helps, here is the SP2004 log:

####################
More than one CPU detected.
You should run two copies of SP2004 simultaneously.
More than one CPU detected.
You should run two copies of SP2004 simultaneously.
Type: Blend - stress CPU and RAM Min: 8 Max: 4096 InPlace: No Mem: 1790 Time: 15
CPU: 3750MHz FSB: 333MHz [340MHz x 11.0 est.]
4/1/2008 11:59 PM Beginning a continuous self-test to check your computer.
Press Stop to end this test.
Test 1, 4000 Lucas-Lehmer iterations of M19922945 using 1024K FFT length.
Test 2, 4000 Lucas-Lehmer iterations of M19922943 using 1024K FFT length.
Test 2, 4000 Lucas-Lehmer iterations of M19922943 using 1024K FFT length.
Test 3, 4000 Lucas-Lehmer iterations of M19374367 using 1024K FFT length.
FATAL ERROR: Rounding was 0.46875, expected less than 0.4
Hardware failure detected, consult stress.txt file.
Torture Test ran 5 minutes 36 seconds - 1 errors, 0 warnings.
Execution halted.

Test 3, 4000 Lucas-Lehmer iterations of M19374367 using 1024K FFT length.
FATAL ERROR: Rounding was 0.5, expected less than 0.4
Hardware failure detected, consult stress.txt file.
Torture Test ran 6 minutes 30 seconds - 1 errors, 0 warnings.
Execution halted.
#####################

Cheers.

bump

[silent-circuit]

bump

Try at 0.4 CPU Voltage. I have a feeling you're expecting too much out of your chip on too few volts. What is your VID?

[Overspeed]

Try at 0.4 CPU Voltage. I have a feeling you're expecting too much out of your chip on too few volts. What is your VID?

My VID is 1.0375. 1.4v will cook my chip, likely destroy it. I have been advised previously not to exceed +.3 in this CPU. This CPU can do 4.2+ easy (a good batch) on Intel boards with air cooling. I guess I'm curious to know if it was the removal of 2 DIMMS that allowed my chip to finally exceed the previous 3.4Ghz barrier or if it was another factor. If so then I might consider chucking my 4x1gb configuration for a 2x2gb. I'd be loosing my D9 Micron, but gaining Ghz on the CPU. As mentioned, SP2004 failed after 5 mins on blend tests. If I can get stable at the new 3.8ghx, I can start from there and work my way to the desired 4.0ghz.

[skyviper]

I've been following this thread over the past 3 weeks I've had my P7N Diamond and it's been extremely helpful. I've been able to get my Q6600 up to 3.6ghz, which is my goal but I can't get my OCCT test to pass. The two problems I run into is either to high of a core temp or a Code 1. My settings are below:

FSB = 1800
Multi = 8
VID=1.325
vCore= +.22500
Nb Volt = 1.45
GTL Ref = 60
FSB Term = 25
Ram = 800, 5-5-5-18 2T
Features = Off

Any tips in addition to what's already been posted would be great.

[dogg]

I've post my results with this cpu some posts below... I was trying settings posted by Yipster, but FSB Term setting offering me max 63 next is auto and then minus numbers... So I've tried 63 and Everest stress test failed after 20 secs.... Now I'm on 3780 MHz /my goal is your startpoint/ since then just one problem with game / test drive unlimited, once crashed with strange behaviour of LCD remaining in vista too, so I decided to restart immediatelly/...
I left many settings on auto ,so it looks like starting point...
There will be somethig about RAMs... I can't set CL4 settings /2x2G/, and your four slots installed could be a reason... I read many posts about problems using all slots... Usually not working properly...
Maybe start where I am and try to uderclock RAMs...

lol


-----------------------------------------
MSI P7N Diamond (780i)
E8400 Penryn on air (Zalman CNPS8700NT)
2x2G Geil Evo One DDR2 CL4 800MHz
2x MSI 8800GT 1GB OC in SLi
2x 150GB WD Raptor
750GB Seagate Barracuda
Power: Coolermaster M series 850W
Logitech G15 Keyboard
Logitech G9 Mouse
Logitech MOMO Racing Wheel
Eizo Flexscan S2031W
Vista 32Bit Ultimate

[JohnNYNJ]

I just finished putting together my P7N Diamond-based system and it has been very impressive and without issue thus far. However, I have a few "open items" that I will outline below...

The set-up:

MSI P7N Diamond, Core 2 Quad Q6600 (VID of 1.1625 in Coretemp ) , 8GB (4 x 2GB) OCZ Platinum Edition DDR2-1000 (2GB to install Vista and then to 8GB after SP1), 2x 8800GTXs in SLI, Auzentech X-Fi Prelude, 2x WD6400s in Raid-0 (OS & Files using onboard controller), 1x WD6400 eSATA (for back-up and using the onboard controller), 3x WD6400s waiting for a home (more on this below), Silverstone Strider ST-1000, Lite-On SATA 4x Blu-Ray Burner, Lian-Li V2000 Plus II and Windows Vista Ultimate 64-bit.

SLI in one blue PCI-E slot and one white PCI-E slot:

In order to save the lone PCI slot for my X-Fi Prelude, I had to use this set-up. They are connected with the one long sli bridge that comes with the board. I started the system off with both of them in the blue slots for comparison.

SLI in Blue PCI-E Slots: 3D Mark Score of 11,194 (no OC)
SLI in Blue and White: 3D Mark Score of 10,223 (no OC)

I played Crysis before and after the change at 1920x1200, settings maxed out and 2x AA and I didn't notice any difference (aside from much better sound from the X-Fi Prelude )

Question: Has anyone got their hands on two of the long bridges and hooked both of them up with the blue and white slot configuration? If so, was there any performance increase? If not, I'm going to be the first .

BIOS issues:

The CPU and Memory are not in the preferred 1:1 ratio. Is there a "quick and dirty" way to go 1:1 on the 1.16beta BIOS? I don't want to play with the voltages until I do all of my homework on oc'n the board. I have a great VID on this G0 Q6600 (1.1625) so I don't want to screw up and fry it before I get to put it on water...

The Plan:

I have 2x 120.3 Swiftech MCR radiators a D-Tek fuzion (with quad nozzle kit), 2 Liang D5 vario pumps and 2 EK full coverage blocks for the 8800GTXs. I'm going to order the Koolance "GPU" block for the NB but what do you guys think I should get for the SB, MOSFETs and PCI-E "sub-SB?" I'm going to transition over to water (modding the v2000 for the two 120.3s) and try to OC to 3.6GHZ.

[malik22]

silent-circuit I see you have this board and the XFX780 which is better and more stable?

[silent-circuit]

silent-circuit I see you have this board and the XFX780 which is better and more stable?

I've managed higher OCs with the XFX board. I'm stable at 3.6Ghz easily, whereas I could only get to 3.4 with the P7N, but I suspect this has more to do with my chip -- X3220 -- than the board, as many others have manage 3.6+ fine with the P7N.

I do believe the P7N will last longer due to its better quality components, but unfortunately I can't afford (and have no real use for) two $250+ boards, so the P7N will be going back to Newegg on Monday.

[AaronYuri]

JohnNYNJ,

This is what Pchas01 did on the MSI forums:




http://forum.msi.com.tw/index.php?topic=115893.0 But he already showed you this.

[allen_845]

I have been reading this board for about 2 weeks now and finally got all of my components to build my new system today. It was good I didn't get them sooner because I had time to reseat the Circu-pipe with Artic silver Ceramique and discovered that MSI had left the applicator tape on the stock thermal material. I was wondering if any one has an opinion on basically flashing the bios right out of the box? After assembling the system I played around with overclocking in bios and got as high as 3.48 ghz (387x9) on the 1.0 bios. I haven't installed windows yet or stress tested it, I just wanted to kind of see the limits of what I could do without causing the screen to just go black (which it did when I tried 400x9). The cpu I am using is a Xeon E3110 (Q805A165). I currently also have 2 X 1gb Crucial Ballistix Tracers installed and will try adding 2 more tomorrow. Also I will be installing Vista sometime tomorrow but will probably lower the settings a little. Any ideas on whether to flash the bios before or after installing windows would be appreciated. Thanks.

_____________________________
MSI P7N Diamond
Xeon E3110 / Zerotherm BTF90
4x1 Gb Crucial Ballistix Tracer
MSI NX8800GTS OC
2x 160gb Deskstars Raid 0
2x Samsung S203N
OCZ GameXtreme 700
Antec P160