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

[loonym]

gamin, I have a hunch that the P09 may cure that 1/2 multi problem. Now all we need to do is find it

[gamin]

honestly, I don't need it right now. my CPU is set as default, with EIST and C1E enabled, so it's working fine. also, my XP32 is loading so faster than before with this 1.1B8. I'm sure they fixed something with the USB 'cos all my problems were always USB related.

here are pics of 9.5 showing up with my bios settings :

idle


100% load

multi 6 and voltage 1.136v on idle, multi 9.5 and voltage 1.200v on full load, so that's fine for me that way...

but I'm pretty sure other people will need/want P09 for testing.

[autumnale]

You know what they need to add to the BIOS? The ability to set the GTL reference for each core. I mean the quads are all two die quads so the two dual cores need to communicate with each other using the front side bus. At least I think that's how it works. And since the two core are different, being able to fine tune the GTL reference for each dual core in the quad would be super nice for overclocking stability.

But I think I'm asking for too much. I'm just happy that at some point in the future, MSI will release an official BIOS with the FSB termination voltage in actual voltages.

[Antong]

hey guys. need some help. i OC'd my B3 to 3.0 and ran a prime 95 for 5 hours with no problems. but when i ran 3dmark06 it stopped and got a blue screen. dont know whats missing. here are the settings i used. can anybody help me?
fsb 1333
clock 667
ram volt 2.10
cpu volt .22500
nb volt 1.425
dimm mem ref 5
cpu gtl ref 30
nb gtl volt 30
fsb term 30
mem timings 5 4 4 13 18 2t

[autumnale]

The following settings are known to support an E8400/E8500 at 4.0ghz and above. Again, the latest BIOS is still finicky when it comes to 45nm chips, so you may not have guaranteed success just yet.

FSB: 1800mhz
RAM: 900mhz

vCore: <<entirely dependent on the OC capabilities of your cpu.>>
vDIMM: <<entirely dependant on your RAM. Set to maximum factory setting.>>
vNB: 1.4
vSB: Auto
FSB Term: 65
All other voltages: AUTO

GTLRef doesn't seem to play as large a role here, and it shouldn't, considering it's concern is more with Quad Core cpus. Compared to the Q6600, the key difference here appears to be FSB Term. 65 (vs 25) seems to be the key for E8400s OCs on this board.

Someone has gotta change the information on the first post in the thread about someone having success with the E8400 with the FSB Term set to 65.

First of all, the FSB Term only goes up to 63 doesn't it? And second of all, any way you slice it 63 would've fried his processor.

If it's 1.xx. Then 63 would be 1.63 volts. That's over the 1.45 volts absolute max for Penryns.

If it's 0.xx added to a base value... It would still be too high.

Someone is going to read that first post and not read the rest and fry his processor.

Better include a warning underneath that.

[autumnale]

I contacted Intel btw, and they told me that the FSB termination voltage isn't packaged on the CPU core. Only the VID information is. That means there is no way for the MSI Tech's answer to be correct. If it is a 0.xx volts that's added onto a base value. That base value must be like 1 volts or 1.2 volts.

It would support the theory that the FSB Term is indeed 1.xx where xx is the integer value you put in the FSB Term.

This is all moot though if MSI releases the next BIOS that fixes this problem.

[loonym]

hey guys. need some help. i OC'd my B3 to 3.0 and ran a prime 95 for 5 hours with no problems. but when i ran 3dmark06 it stopped and got a blue screen. dont know whats missing. here are the settings i used. can anybody help me?
fsb 1333
clock 667
ram volt 2.10
cpu volt .22500
nb volt 1.425
dimm mem ref 5
cpu gtl ref 30
nb gtl volt 30
fsb term 30
mem timings 5 4 4 13 18 2t

It very well could have been a vid card or driver issue that blue screened you. Prime is not an indicator of graphics stability.

[loonym]

First of all, the FSB Term only goes up to 63 doesn't it? And second of all, any way you slice it 63 would've fried his processor.

http://forums.hardwarezone.com.sg/sh....php?t=1886633
http://www.xtremesystems.org/forums/...&postcount=134
http://www.xtremesystems.org/forums/...&postcount=146
His cpu is very much unfried and he's still benching with it in new boards he tests I believe.

[loonym]

This is all moot though if MSI releases the next BIOS that fixes this problem.

See post 572

[theYipster]

First post updated to announce new BIOS release. New suggested settings for the Q6600 are coming once I get a chance to play around with it (likely this weekend.) Anyone else with 1.1B8 experience? Please share results!

[theYipster]

Antong,

As Loonym suggested, it could be a driver issue. However, let's check a few things...

Couple of questions: How are you running SLI? (both in blue slots or one in blue and one in white?) If one is in the white slot, you may want to give your SB voltage a bit of a boost, as I believe the white slot is handled via the SB. Try increments of .05v and see if that improves things. Next, how did you arrive at raising the NB GTL voltage? Have you tried to find stability with that on auto? GTL voltages can be finicky. Setting it too high may cause the problem.

Lastly, you want to be sure that your testing thoroughly. Prime95 is a good CPU test, but other tests are better for NB and memory. When running Prime95, be sure to set round-off error checking to on (in the menu,) and be sure you are running the Small FFT test. Also, many don't consider a Prime test worthwhile until it runs for at least 8 hours, while others don't consider anything under 24 hours as being able to prove stability.

For NB and RAM, you'll want to get Orthos 2.0 and run a custom NB only test. If it fails within an hour, you know you have some work to do adjusting your NB voltage settings.

Good luck.

[pacific vortex]

the 1.18 is showing promise...other than a glitch with the sata dvd that it wouldnt recognize at first...no usb probs or anything..just finished first 30 mins of occt at 3.825gignow for a long nite...fsb terminator voltage seems to be capped at 1.315 v

[overheat13]

for information i got one fsb holes on this board that come with 1:1 memory ratio for 3.2ghz = 1422 with 800mhz. but still not stable at the moment. just got 17 minute. i dont know where's my wrong setting. for the vcore i just using 0.012500v, ram using 2.00v, n nb voltage using 1.45, with 65 gtl and 25 fsb terminator. please if somebody having a good one for this fsb.

[AaronYuri]

After clearing the CMOS, go into the BIOS, do not change anything, save and exit, then boot to windows. This allows the BIOS to build DMI pool data. If you don't do this you will run into problems.

[zillmc]

Hey all, I hope everyone is doing well!

Now that my comp has been running stable for a month and with the release of the new beta bios for the P7N, I'm going to embark on my overclocking adventures. As I'm still very new to overclocking I thought i'd post here to get some input from you guys. Please see my system specs in my signature for details.

I'd like to start off with a modest overclock for my X3350 @ 2.66ghz to around 3.2ghz with 400x8 is my guess. Can you guys recommend which version of bios I should be using as well P07 or 1.18?

Additionally, what settings should I start out with for DRAM freq, memory voltage, vcore, fsb term, nb/sb vcore, memory timings, etc. Any information would be great!

[loonym]

I'm kind of liking the 118 for the voltage inputs. If your going to run 400fsb you may as well go 1:1 with the mem. You may find that at 800 you can tighten up the timings on the reapers with 2.1v

[zillmc]

So I flashed to 1.18B and have started trying to OC without any success. For some reason when I change the FSB from 1333 to 1600 and mem to 800 and save and exit it wont reboot. It will just hang and the monitor lights will stay amber color.

I'll have to reset/power on and off a few times before I get it to POST and it says "Last overclock failed loading default settings.."

Another thing to note is that all my voltage settings are grayed out and I cannot modify them. Any ideas?

[overheat13]

Finally i got my q6600 stable at 3.2ghz with shipping bios! it seems the important things is the Fsb termination. When i try auto, 10, 15, 20, and 25 my prime test will hang before one hour. now here's my setting that i got stable :

fsb : 1422 = 3.2ghz
memory ratio : 800mhz

vcore : 0.150000v
ddr : 2.1v (5-5-5-15/2T)
nb : 1.5v
gpu gtl : 65
fsb term: 35

idle 28c / under load 62c

with this setting i got more than 24 hours running prime without any error.
here's my spec:

q6600 GO 2.4ghz = 3.2ghz
p7n diamond
mushkin ddr2 EM pc6400 800mhz 4x1gb
vga zotac 8800gt amp edition x2 in sli mode = around 17380 in 3dmark06
case shark by thermaltake
2x 90mm fan on side panel
using Thermalright 120 ultra extreme with 2 fan 120mm from thermaltake with high rpm (made me cannot sleep tight with that sound!!!)
HDD 320 gb Seagate 7200rpm Sata II
HDD 120 gb Seagate 7200rpm Sata II
plus one antec spot cool for my northbrigde

thats all. for me enough to get 3.2ghz since i cannot get stable more than that by using shipping bios. Thx for any information that can help my oc! GOOD LUCK guyzzz..

[hump]

1.1B8

2000/1000 x6

1 hour OCCT stable so far

vcore : 1.18 load reported in Everest (I think I set it to .25000 in bios. VID is 0.9)
ddr : 2.2v (5-5-5-15-24)
nb : 1.5v
fsb term: 1.306v
sb : 1.8v (maybe it should be auto)

Everest just locked up while viewing Everest CPUID. End tasked it and reopened it and it's running again but OCCT hasn't errored yet... weird

*EDIT* The wierdness that I get could be from using the same install of windows in 3 different computers all with different nvidia (NF4, 680i & 780i) chipsets lol... I really need to reinstall.

[hump]

I let OCCT run for 7.5 hours. Shut it down and started a 3DMark 06 run and in the first test I was getting 15 to 20 (85fps vs 100+) frames less than I was at 1600/800 x8 (3.6GHZ). I would have thought even at 3GHZ with a 2000/1000 bus, it would still have better fps than 3.6GHZ @ 1600/800. I think it's just because my CPU is the bottleneck with the 8800GT SLI setup...?

I thought that it may be a memory hole so I tried 2010/1005, saved and exited bios and the pc just shut off. Took hitting the clear cmos pushing the power button more than 10 times to get it to come back on. The first couple of times nothing would light or spin, then the LEDs and fans would power on but the lights on the ballistix wouldn't come on, then lights, fans and ballistix lights would come on but no POST... and then it gave me the message to hit F1 to enter setup to fix unstable settings which tells me that the clear cmos isn't always working?

It did something similar with the P07 and i ended up reseating the CPU to get it to boot again...

I will update when I have the time to do more playing.

[hump]

After having to reseat the CPU again to get it to POST, I'm sticking with 450x9 until a new BIOS comes out. That gives me 1900/900 @ 4.05GHZ stable, and I can reboot w/o having to reseat the CPU.

It's stable at 2000/1000 all day, as long as I don't reboot... that makes me sad.

[dogg]

some results with dual core....
it's about 4 weeks i'm running settings as I posted last time..
There was some minor issues like hangouts during games/ very occasionally / and driver crashes but nothing fatal... I changed graphic drivers several times and now running well...
So probably driver issues.. Sometimes I'm not sure about what caused crash, because on bad oc I got more driver crashes..
But as I wrote, now for 4 weeks fine.. Thanks loonym for so much help with it...
I think it could run 4.05 GHz as well, but I like temperatures with these settings and summer is comming too

This maybe help someone ...

BIOS P07
FSB 430 x 9 /3870/
RAM 860 /ratio 1:1/
CPU V +2250
NB V 1.375
RAM V 2.1
RAM Timing on AUTO /5-6-6-20 T2/

Tried FSB Term voltage on +63 and no frying of my processor at all...
Even, I think, no changes on temperatures... No stability improvement as well.
I got a feeling it's not working with dualcore..

Didn't try 118 Bios yet... Someone with that BIOS and E8400/8500?

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
Coolermaster Realpower M 850W
Logitech G15 Keyboard
Logitech G9 Mouse
Logitech MOMO Racing Wheel
Eizo Flexscan S2031W
Vista 32Bit Ultimate

[Kano1977]

I've said from the start that constructive dialogue and reporting known and reproduceable bugs and quirks (not simply complaining) with the support people at MSI would produce positive results. At the same time we have to be patient because they have scores of other products to support.

I am new to this overclocking thing.. So i tried your setting to get to 3.6 with my 6600 everything seems stable so far .. Now i am using a zalamn cooler my idle tempts are 109 f and full load 149 150. Are these tempts to high for a overclock of 3.6 ?

[Kano1977]

Also i have my vcore set to 0.225. That seems high to me should i worry about burning out the cpu with the volts being so high ?

[Kano1977]

Here are bios screens of my settings for 3.6g (9x400). For higher clocks the only ones I change are nb volts to 1.525 and vcore increase depending what frequency I'm trying to run. The increase in pci-e freq and expander voltage are only for me to help stabilize my vid card oc.

These are the settings i tried