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

[Mannekino]

loonym,

With regard to your "regular" voltage settings, might they work for me with my E8500 overlock issues? Please see a my two posts on page 28 for more details. And how do you get your "System Temperature" so low? Mine if 44°C when idle. And I believe my case is pretty nice with regard to cooling.

- Mannekino

[Cal]

Cal,

If the bootblock is still intact a recovery is possible.

Was your system overclocked before the flash? You should always make sure your system is stable before flashing...

Cheers going to pick up a ps2 keyboard today and give it a try before i rma

system was running stock and i thought stable but i was mistaken

[loonym]

The small office where the pc's are is the location of a 10,000btu window ac unit so ambient is very good. It blows directly through the pc area before the air moves out into the surrounding rooms. Also, The TRUE benefits a great deal from high cfm fans.

[loonym]

The pci-e settings have nothing to do with the board or cpu oc, I use those for my vid oc's. As far as actually explaining voltages like dimm mem ref, that's a litle out of my league but I was always taught (through forums like this) that a good setting for that value is 50% of vdimm. I have no idea if it helps me stabilize but it doesn't hurt All my settings are settled on after experimenting with my own various setups after periods of trial and error and are examples for reference only. I would not recommend any of it for anyone else but I do recommend patience and experimenting. I have killed a lot of gear so take anything I post with an x-large grain of whatever.

[Grimspoon]

The P7N Diamond is high on my list of potential motherboards for my next build, can I expect to get 8gb of ram working on this motherboard (Patriot Extreme Viper)?

Also, if I were to buy this board along with a GTX 280, how likely am I to experience video corruption? These are the two main issues that have prevented me from going ahead with my next build, and I'd like some first hand responses from experienced P7N users, so your input would be greatly appreciated.

Here is what I'm hoping to put together:

Intel Core 2 Duo E8500
EVGA E-GEFORCE GTX 280 Ftw
Patriot Extreme Viper 4GB 2X2GB PC2-8500 DDR2-1066 (x2)
Western Digital Caviar SE16 640GB (x2)
Bgears B-ENSPIRER 7.1 DDL DTS
Noctua NH-U12P LGA775 AM2 Heatsink
Noctua NF-P12-1300 120MM Ultra Quiet Cooling Fan
Corsair TX750W 750W
LG GGC-H20L BLU-RAY HD-DVD Reader BD-ROM
Antec P182 ATX Black Mid Tower
Logitech G5 Laser Gaming Mouse
Logitech Internet 350 USB Keyboard Black

Chances are if I feel like the video corruption issue is going to be a huge problem I'll probably go for a P45, but I'd prefer to buy the P7N and keep my SLI options open...

[loonym]

The P7N Diamond is high on my list of potential motherboards for my next build, can I expect to get 8gb of ram working on this motherboard (Patriot Extreme Viper)?

Also, if I were to buy this board along with a GTX 280, how likely am I to experience video corruption? These are the two main issues that have prevented me from going ahead with my next build, and I'd like some first hand responses from experienced P7N users, so your input would be greatly appreciated.




Chances are if I feel like the video corruption issue is going to be a huge problem I'll probably go for a P45, but I'd prefer to buy the P7N and keep my SLI options open...

I never experienced the video corruption but I see a lot of people saying that the latest beta bios has solved their issues. There are a couple good examples of people running 8gb ram earlier in this thread although I've never needed to try it. I certainly think it can be done with the right voltages to nb, etc but I wouldn't expect to run much over ddr2-800 speed.

[AaronYuri]

Ditto there, loonym. Don't even think X48 would get 8GB over 800MHz.

I experienced the corruption issue, and BIOS v1.2B1 solved it for me.

Ordered a Q9450 today, so I'll post my findings once I toy with it.

[Fiskbit]

I just set up my computer today and have been having stability problems. Everything's running at stock. I flashed the BIOS with v1.21 and it seemed to fix my stability issues, but I just got an IRQL_NOT_LESS_OR_EQUAL crash, so something's still wrong. I'm no longer able to boot up the computer; it fails with "A disk read error occurred. Press Ctrl+Alt+Del to restart." Attempting to get into BIOS settings yields the same result.

In addition to the stuff in my sig, I've got a SATA Samsung DVD-RW and an older IDE DVD-RW. My CPU's VID is 1.225. Earlier, I tried removing a stick of RAM to see if that'd help, but the crashes continued.

Any suggestions? I'd really appreciate it.

[loonym]

That error sounds like your boot order is wrong or maybe something wrong with your hdd like mbr failed. You can't enter bios at all at this point? Have you cleared the cmos?

[Fiskbit]

I cleared CMOS and got into the BIOS. The boot order is currently:

1. Floppy
2. WD Caviar 640GB
3. WD Caviar 640GB
4. Seagate 750GB
5. DVD-RW
6. DVD-RW

The two WD drives are in RAID 0. After getting the BIOS settings back to how they were, the computer booted up just fine!

This is the 4th or 5th time it's crashed today, though, so something's unstable. What do you recommend I change to try to stabilize it?

I apologize for my newness; I'm entirely new to overclocking. I'm looking forward to learning how all of this works. Thanks.

Edit: Just rebooted again out of the blue. Happened very quickly and gave no error.

Edit 2: I tried a 75 MHz overclock without touching anything else but RAM (running 1:1) and things are running stable so far. Perhaps it will stay that way. Here's hoping.

Edit 3: Still can't get this thing stable. I'm running at 2.55 GHz now and don't know what to do to get it to stop crashing. Just been slowly upping the FSB, but I can't keep doing that because the core temperatures are rising (currently idling at 39 C).

[loonym]

I cleared CMOS and got into the BIOS. The boot order is currently:

1. Floppy
2. WD Caviar 640GB
3. WD Caviar 640GB
4. Seagate 750GB
5. DVD-RW
6. DVD-RW

The two WD drives are in RAID 0. After getting the BIOS settings back to how they were, the computer booted up just fine!

This is the 4th or 5th time it's crashed today, though, so something's unstable. What do you recommend I change to try to stabilize it?

I apologize for my newness; I'm entirely new to overclocking. I'm looking forward to learning how all of this works. Thanks.

Edit: Just rebooted again out of the blue. Happened very quickly and gave no error.

Edit 2: I tried a 75 MHz overclock without touching anything else but RAM (running 1:1) and things are running stable so far. Perhaps it will stay that way. Here's hoping.

Edit 3: Still can't get this thing stable. I'm running at 2.55 GHz now and don't know what to do to get it to stop crashing. Just been slowly upping the FSB, but I can't keep doing that because the core temperatures are rising (currently idling at 39 C).

I think the best advice to you is to slow down a little and read a lot. Overclocking, to me, is more than just grabbing a set of good looking numbers that someone threw up on a random website screenshot. The numbers I decide on for my hardware are usually arrived at after many hours (days,weeks?) of playing. Most times this includes numerous flashes to find 'my' best bios, dozens of crashes, freezes and bios resets, a few reinstalls (trueimage is my friend), hardware swaps, and the settings hopefully evolve as I dream (nightmares?) up something new to try. When I see someone saying 'my system is crashing what do I do?' the first thing I look at is ram, then power supply, and cooling. If those are satisfactory the next is timings. But 99% of the time the answer to oc problems is voltage. Naturally more voltage=greater cooling needs. If I can't get stable I generally don't have enough voltage somewhere for a given oc. Then it becomes a process of elimination, trial and error, and most of all patience.

[gamin]

what about an USB devices problem ??? have you tried removing all your USB devices if you have any ? I know, this can be hard if you have both mouse and keyboard on USB...

@ Grimspoon : for now, I don't regret my choices (see my sig for full spec). looks like you're about to build a killer rig (bios 1.2b1 fixed the video corruption problem I always had, and also fix the .5 multiplier availablility for my E8500)

[Fiskbit]

loonym: Yes, I understand that; all hardware is different, and I don't expect overclocking to be a breeze. I'll be holding off on trying to achieve a high clock until I have more adequate cooling, and I realize that even then it will be very slow as I tweak lots of stuff and run stress tests for long periods of time. The thing that concerned me here was that the board isn't running at stock, but I recall that other people earlier in the thread had that issue. I'm getting a lot closer to stability now, with a higher clock. Looking good.

My power supply shouldn't be an issue. Cooling is also fine; temperatures are all safe. My RAM is currently at 566 (1:1 with FSB). Its timings are 5-5-5-15. As for voltage, I'm avoiding increasing it, as I'm running stock cooling. My current settings may be working, as Prime95 has been running for a while without fail, but it's too early to tell. Should I be tightening the timings on my RAM at all?

Thanks much, loonym. Your posts regarding success with this board in this thread eased the worries I developed after ordering the board and reading about some of the early troubles people came across.

gamin: My keyboard isn't USB and my mouse, while USB, is hooked up through an adapter. I've had a USB cam hooked up at times, but it didn't seem to have any negative effect (though the extra beep when powering up concerned me for a bit until I realized it was the USB!). I'm actually a bit surprised that I haven't come across several of the problems others have seen, such as SATA issues. The machine's running a lot more stable now, so I'm hoping I don't run into any more problems. Aside from not running at stock, I'm really, really liking what I've seen of this board so far.

[dogg]

I had overclocked my e8400 to 3.825. I set my fsb to 1700, dram to 850 cpu v to0.2, dramv to 2.1 nbv to 1.350 and sbv too 1.550. It ran stable for a while I ran prime95 for 10 hours no problems. My temps were a little high for my liking so I bought a zerotherm nirvana and an asus dx and about a couple of weeks later I decided to run at stock to check temps and the what not and when windows booted up I got the bsod and it will happen anytime I boot into windows at stock speeds not with the voltages I used to oc. When I went into hardware moniter in bios I saw that my effciency was ranging between 19 and 25%. I swaped out my psu and same thing could it be my mobo? Please help someone I dont know what to do.

I've had the same situation after some time when I was running unchanged OC and then decide to try higher... Answer was somewhere else than in bios.. I'm not sure if it's the same issue but it looks like.. My crashes started after vista startup when background programs was starting but just when I changed my OC /up or down or default/... So I had feeling it's software based problem... That was some profile I've set while I was not changing OC so I did't see anything wrong.. Profile was set in nVidia control panel, I used it just for fan speed setting but it saved my OC as well and that settings are running after system startup so it probably trying to change bios OC and crashing /with new vid drivers is this nVidia OC utility again as nTune and disappeared from control panel, I've read at nVidia site.../ So I set my OC to the same as were in that profile, started up system, deleted profile and everything is nice and shiny now....


panheimer
Really no issues with G15+G9 and other usb devices after 1.2B1 flashed.... I'm trying to keep chipset drivers up to date and it's working for me /always using those from nVidia site /

grimspoon
I was surprised that video corruption was bios issue / believed it's driver problem / but with 1.2B1 it's solved..

[Fiskbit]

Ultimately, I think my issue may have been related to SATA rather than anything else. After running Prime95 for hours, I had no troubles, but as soon as I was doing general use again, I ran into problems. I installed the SATA drivers for the 790i after seeing in a thread somewhere that they work and have had no more issues. For anyone having trouble with SATA, this is probably the way to go.

[AaronYuri]

New beta out: http://forum.msi.com.tw/index.php?to...3750#msg863750

Dated 25/06/2008. As usual post feedback in the thread.

P.S Got my Q9450 today. Currently running 400 x 8 @ 1.2v VCore. I'll update later after I push much further.

[BababooeyHTJ]

I have been having some hard drive issues. I increased my south bridge voltage to 1.55 seems to have helped and I get my best results in everest disk bench at that voltage. I am using 2x WD caviars 3gb/s 16mb cache 7200rpm on a 64kb stripe (I should have used 128) and I get a really high average read access (16.88ms). My other scores are good Liner read (begin)= 150.4MB/s linear read (middle)=132.4 MB/s Linear read (end) 73.8 MB/s random read 113.9MB/s Buffered read 316.5MB/s What could be causing my high average read access?

[loonym]

There's a new beta bios, you can d/l the 1.22 beta at msi forum. http://forum.msi.com.tw/index.php?topic=115073.0 Registration is required I believe to d/l

[Fiskbit]

To those of you running SLI, what did you have to do to get it enabled? I have two 8800 GTS cards and have SLI set to enabled in the NVidia control panel, but I see almost no change in 3DMark06 score. When I enable "Show SLI Visual Indicators", there are no onscreen indicators in anything I've tried (Half Life 2, Crysis, 3DMark06). I've looked all around and can't figure out what may be wrong; is there a step I missed?

[loonym]

Sometimes enabling sli requires a second install of the vid drivers.

[Fiskbit]

I've uninstalled and reinstalled a few times, but it doesn't help at all. I'm using the newest Geforce 8-series drivers.

[AaronYuri]

I don't have a change log for the BIOS. I have no idea what has been done to it. If anyone has an idea, it will be Jack. He will post it here: http://forum.msi.com.tw/index.php?topic=115073.120

Nope, at the moment I don't have any info.

[XavierMaxx]

I've had this board for about three weeks and I felt like sharing my experiences here as I found this thread invaluable for both research and troubleshooting. This board was attractive for what I wanted and I decided to take a bit of a risky plunge when I purchased it - I come from a long line of AMD-based systems and I was never big on SLI. I've of course been a guest/lurker at XtremeSystems for quite some time and I'm a member of many other similar forums, and likewise I have a good deal of experience under my belt. My current system configuration is in my signature for reference. I only hope that my issues and solutions with this maturing board will be useful to others, via my elaboration.

I had two major problems with this build that I've been tackling for several weeks now, and I've only found a solution to one of them. The first problem is characterized by the inability to get the system at all stable past about, well, what I have my system at currently. Usually the system simply freezes, usually a few minutes into OCCT, although if I jack up the FSB voltage and tweak the GTLs I can get it stable up to 30 minutes of OCCT. I generally attempt this at a FSB of 1422 (9x355.5 = 3.2 GHz), and regardless of voltages or other settings (it does pass Memtest) it will freeze in that manner, although again a relatively high NB (1.4-1.45) and FSB voltage (1.313) will get it progressively more stable. I imagine if the board had higher FSB voltage settings I might be okay, but I still find it a bit perplexing considering the ease of 3.1 GHz stability.

The second problem was that the system would OCCT fine from stock to 3.1 without a problem, but games would eventually crash. The games exhibited behavior such as orange checkerboards, VPU recoveries (brief black screen), freezing, speeding up of visuals after a pause, CTD, etc. The higher the OC, the faster this happened; I tried switching the PSU and memory before finally in a last ditch effort I wedged a decent fan over the NB. Everything is quite stable, now, so I conclude the NB was overheating, although why it would do that even on stock is beyond me. My best theory at the moment is the motherboard design coupled with my CPU cooler configuration somehow trapped a lot of hot air in that region.

This motherboard also has some other small issues - the known USB keyboard issue - that I won't elaborate on too much. One of them is that the 2nd LAN's USB ports don't register with the BIOS properly (if I plug in a flash drive to one, it won't be listed/detected on POST as a storage device) although they work fine in the OS. The second is a proclivity for this motherboard to come up blank and need a BIOS reset after unplugging the power cable and moving around hardware - this did not happen on my PSU switch, but it did on memory switching and slot changing, X-Fi slot changing, and front USB hook-up. These were particularly noticeable in BIOS v1.1 but I still have those issues in v1.2b1 to a lesser degree.

I would appreciate any assistance on the FSB problem, and I hope my solution for the gaming crashes will be of help to some users. It's possible my board has specific issues or my configuration is unique, but I just don't like giving in to RMA/replacement. I took some time testing and organizing my findings and wanted the information out there. This thread seemed the best place to do that, so pardon the length of my post. XS is the only forum I know that can fully appreciate the crazy tinkering of hardware pioneers.

[Fiskbit]

I've been doing various tests in 3DMark06 and have managed to get scores from 12,000-14,000 with SLI enabled in different modes. Forcing software to use only a single card, I get around 12,000. I've also not noticed any improvement in frame rates in Crysis or Half Life 2. From what I've been reading, I should be seeing more improvement; is this normal, or is something wrong with my system (perhaps a CPU bottleneck or some other issue)?

Both 8800 GTS 512 cards are running at 650/972/1625.

[XavierMaxx]

I don't have my system set up in SLI yet, Fiskbit, so I'm sorry I can't help you, but I am getting my second card this week and I will report my experiences in 3DMark06.

Also I'd like to update the thread on my situation as I believe I have finally made progress on the FSB issue. If you recall from above, I mentioned that 9x343 was happily stable while 9x355.5 wouldn't budge for the most part. Well after playing around with some of the memory voltages and other things the other day, I noticed more stability at 3.2. I also noticed that while I had to set SLI-memory to Expert for 3.1 to be stable, it had to be Disabled for 3.2 to even POST. I came to the conclusion earlier tonight that the motherboard's auto values for my RAM's TRC was dependent on FSB such that the auto values for Expert were too low sometimes while auto values for Disabled were too low in other times.

I had written down all the values - literally all - that CPU-z and Everest reported for the memory timings, and compared the stable 3.1 to various other configurations (3.2, 3.4, 3.6, etc.). I toggled SLI from Expert to Disabled as well and checked those values. What I noticed was that the memory was only stable at a TRC of 48T (45T is SPD according to CPU-z) but that the system was bootable at 32T (which is what SLI Disabled defaulted to at 355.5 FSB) and not stable at all. I also noticed that at FSB 400, for example, SLI-memory needed to be enabled but had a value of 36T, but SLI enabled at FSB 355.5 was unknown (no POST) and likely sub-32. This was regardless of what I set for memory speed - 1066, 1000, 800, etc.

So I went back to 355.5 and booted with SLI-memory Disabled, but then used Memset to change the TRC to 48T in Windows. Voila - OCCT for an hour just as a probing attempt went without a hitch, where setting any value to RAM in BIOS right before it would result in a crash within 5 min. Even top voltages managed only 20-30 min on lucky runs. So I'm pretty confident that I have at least narrowed down the issue.

The quick report is this: users having instability in games but not OCCT/Prime should consider active cooling of the northbridge, even on stock settings. Users with problems getting past 3.0-3.1 GHz on a quad, or past 350 FSB, should go with bootable/POST-capable BIOS settings for memory and then use memset to raise TRC in Windows. FSB Term and memory voltage of course help, as is well-known, but these two particular problems seem to have specific solutions. Hope that helps somebody and I'll be back to diagnose my SLI problems that I'm sure will crop up.