definately board or bios problems because with 2Gb sticks I've problems going above 820-830mhz. will let msi know about this.
I have no problem at all pushing my G.Skill 4GBPQ to >1000MHz on this board. :rolleyes:
Gotta say this board kicks loads of butt with this quad and it's a huge step up from my P965 board. Managed to hit DDR2-1176 5-5-5-15 with relatively loose subtimings with only 2.3v with my D9GKX. The kicker though was 5-5-5-15 timings, 1.8v, and 1032MHz 24/7 stable :D Also does 3-3-3-9 2.3v 826MHz 24/7 stable, when I get back going to pump 2.5+v and see how this board will really fly. The limited quad testing I've done has shown 500FSB stable with moderate to high motherboard voltages but CPU voltages themselves are shockingly low. Going on vacation tomorrow so I won't have a chance to post any results, but for $90 or so this board is definitely a good buy.
v1.6
I tried 1.6, I can go higher with that .. like 1080MHz on the memory but on very lazy latencies .. like 5-7-7-21. Compared to 1:1 has no advantages except that you get higher nr on MHz, but on real memory bandwidth you get the same or even lower transfer rates due to high latency.
I have had great luck with my MSI p6 platinum 650i. It cranked on overclocking my Q6600 to 3.6 on air no problemo, but i want a 780i board now...
Got another 4GB kit. This time it's Patriot Signature 4GB kit. Same problem. Can't pass FSB 420. Tried BIOS 1.6 too. Seems like compatibility problem with 2GB sticks.
I cannot get it too run with the 1:1.25 divider (947). It wont even boot into windows. Runs fine in the 1:1.2 divider (909). I don't think it's my memory, G.Skill 4GBPQ, because I had no problems reaching 1000 at 1:1 divider. May be a 1:1.25 divider issue with the board. Using 1.1 bios. Anybody experiencing the same?
I dont know if i should RMA my own Neo2-FR ....
CPU: Q6600
1. It has never shown CPU Fan rpms being more than 1100 (AFAIK shoudl be 2000-2700) ... I get 77C under orthos load at stock 2.4 and 84 @ 3Ghz .... Either is misreading both CPU fan rpms and temperatures or is not controlling the CPUfan at all and indeed temps raise that high ... smartfan is not working anyways no matter what settings
2. Lost two times the 16x link width of the PCIE16x to 1x and then revived again, have no clue how and why
3. all bioses above v1.3 will give me as starting VCORE setting the 1.30v (instead of the proper 1.25v). 1.3v will be the lowest option in the vcore menu settings .... (and the measured voltage with this setting is 1.272)
4. bios 1.76 wont save the settings at all always post up with "bad checksum error Press F1 to enter setup or F2 to load default values and continue" message. No matter what i do (I think this happened with 1.6 too). Ofcourse did complete CMOS clearing, no power etc
1. Find out what CoreTemp reports about your temperatures, should't have problems reporting the right ones.
2. Could be because of bios flashing problems.
3. They are set to a 1.3000v default because the vdroop is bigger, you should normally get 1.275 (default) when set to 1.300v . For example, in bios 1.1 you get 1.3000v in idle when you set 1.2750 in bios. From 1.3 -> it's the other way around.
4. Looks to me like you didn't use afud 408 and you used 412 or 418 that sometimes writes the bios with errors and they don't destroy cmos data.
1. Did you enable Smartfan?
2. Not sure what you did.
3. All bios after 1.3 will set the VID of the CPU to the lowest Vcore.If the VID of your Q6600 is 1.3V then the lowest Vcore setting is 1.3V
4. Try to reflash the bios again. I had this problem last time and a reflash will fix the problem. Btw 1.67 beta7 is around.
I contacted MSI technical support and they told me they will investigate about compatibility and overclocking 2GB memory sticks. Here is a reply i got :
Can anyone check their board and let me know the MB'ERP number? Thank you.Quote:
Dear SIr/Madam,
Thanks for contacting MSI Technical Support again.
Please kindly provide us your MB'ERP number(refer to the below picture) and we will copy your problem.If we can copy this issue and we will fix it ASAP.If we release the new BIOS and we will contact you ASAP.
Thanks for your cooperation in advance.
Best regards
MSI Technical Support Team
http://support.msi.com.tw/images/up_...8474437_up.JPG
Mine has 030, i have that same memory and can only run it at 1:1.2 divider. Thanks for informing MSI a bios fix will certainly help us.:up:
Sorry m8, when i got the settings on my sig p95 stable and can pass 32m superpi i dismantled my system to mod my case to fit in another mcr360 rad but my ram is running at 1000mhz at 2.0v i think. I'll try for the max i can get when i put it back on.
wyemarn I have this memory, 2 kits and if I use only one kit (2x1GB) I can go way over 1000MHz on 2.0v, I was at 1080MHz 5-5-5-16 2.0v and there was room for more but didn't push it.
This is the memory I'm talking about:
http://happy.ap.teacup.com/occ2/3.html
i have this board since 1 month now , using 1.6 official bios (flashed today with 1.7 beta 7 bios )
Using 2x1GB OCZ XTC Titanium EPP @ 1002Mhz 4-5-4-12 2T ( 6-5-4-5)
+ 1 Micron D9 Based 800Mhz memory @ same speed total 3GB ;]
Xeon 3060 can do easy 7x500Mhz @ 1.478V = 1.465 Real
Runnin 24/7 @ 9x401 ~3609Mhz + EIST Enabled ,
Cooled by Scythe Ninja Rev.B
Great board.
Will post pics soon.
I'm running Biosversion 1.63B now for a couple of weeks and this beta seems to be quite stable. The only thing witch is bothering me is the fact that you can't undervolt the cpu. are there any other biosversions which are offering the feature of undervolting?
Hi tty, Thanks for the reply. I've used that memory before. I can push that beyond 1000mhz too. Now with 2GBx2, I cant go beyond 840mhz. Are you the writer of that review?
If I'm not mistaken, BIOS 1.1, 1.2 and 1.3 allow the user to undervolt 0.05V below VID. So if your VID is 1.35V, you set the lowest Vcore to be 1.3V.Quote:
I'm running Biosversion 1.63B now for a couple of weeks and this beta seems to be quite stable. The only thing witch is bothering me is the fact that you can't undervolt the cpu. are there any other biosversions which are offering the feature of undervolting?