• Some users have recently had their accounts hijacked. It seems that the now defunct EVGA forums might have compromised your password there and seems many are using the same PW here. We would suggest you UPDATE YOUR PASSWORD and TURN ON 2FA for your account here to further secure it. None of the compromised accounts had 2FA turned on.
    Once you have enabled 2FA, your account will be updated soon to show a badge, letting other members know that you use 2FA to protect your account. This should be beneficial for everyone that uses FSFT.

Help me raise my overclock

P

pcfountain

Weaksauce
Joined
Aug 4, 2004
Messages
101
MSI K8N Neo4 Platinum (non-SLI)
A64 3000+ Venice
Arctic Cooling Freezer64
2x512MB Corsair XMS
PowerColor X800XL 256MB
OCZ Modstream 520W Power
Lian-Li PC-67

I've read all the guides here which have been very helpful and I've been struggling with this for about a week. Here's the situation: the highest settings I can boot at are 4x240HTT, 9xCPU for 2160Mhz. RAM is at 200Mhz Divider, 2.5-3-3-7 Timings. It does 10 passes of test #5 in Memtest86, no errors. All of this on stock voltage. If I try to push it to even 241HTT in the BIOS, the machine will not post.

However, once I boot into windows, I use ClockGen to slowly ramp up the HTT while running Prime95 + Rthdribl to stress the system. I got it all the way up to 267HTT x9 = 2403Mhz with the RAM at those same settings. I have now been running Prime95 + Rthdribl stable for about 9 hours at this setting. Again, I am on stock voltage, my highest CPU temp so far is 41c, system temp 35c.

How can I get my machine to post and boot at higher settings? Some folks have suggested upping the RAM voltage, but that has not worked for me. I can take a working setting, and if all I do is change my RAM voltage, it will not post. Dropping down the RAM divider doesn't seem to help either. I tried 241HTT, and started dropping the RAM divider with timings at 3-4-4-8 ... I had to drop it all the way to 100Mhz before it would boot. It would seem that my RAM is the limiting factor here, but why can I get an extra 27Mhz out of it in Windows at the same timings and stock voltage?

:confused: Frustrating .. hoping someone can help.
 
I flashed the latest BIOS when I first set up the system about 2 weeks ago. Haven't looked since then, but MSI's live update utility makes me want to throw up. So I'll check but sort of doubt this is the problem.
 
Well Prime95 failed after 17 hours. So maybe I pushed it a bit too far. Still, by dropping the RAM Divider a bit, it seems like I should be able to reach my goal of 2.4Ghz on this puppy. Maybe a screenshot will inspire somebody ...
oc1.jpg
 
OK, something is really wrong here. I am starting to contemplate doing some very bad things to my computer. Nothing was working, so I went back to Eclipse's OC guide and started over from square one. This setting boots just fine:

HT link multi 2x, CPU multi 4x, RAM divider 100Mhz, HTT 240Mhz

If I try even one more Mhz - 241 HTT in BIOS, it will not boot. However, I can boot into Windows at 240 HTT and use ClockGen to go way, way, way up. I stopped trying at 290 ... because it's a lot more than I need to reach my goal of 2.4Ghz.

I checked some of the things mentioned in the OC guide ... diabled SATA because I'm not using it. PCIE bus is set to 100Mhz, don't even see an option for PCI but ClockGen tells me it's at 33.34Mhz. Tried other HT link multi, 1x, 2.5x, and 3x, no dice there. Also tried raising vcore and nforce voltages ... this board seems real touchy. As soon as I change any voltage from stock, it won't boot.

HELP!!! I know there is a good overclock hiding in this setup, just waiting to be unleashed!! Why can I go to big speeds in Windows/ClockGen but not boot them in BIOS??
 
Its possible ClockGen isnt actually giving those speeds right? did you benchmark before and after to see if it works?

And as many people will tell you (and I will confirm), a platform that cant prime95 for at least 24 hours straight is too unstable to be worth using.
 
Well, by doing some searching on other forums, I'm seeing many alleged reports of a BIOS bug with the K8N Neo4 Platinum and Winchester CPUs. Apparently a lot of people can't post higher than 220 or 225 HTT. I wasn't able to find similar reports on Venice CPUs, but I have to venture a guess that this is exactly the problem I'm having. Of course there is no official word from MSI but a lot of enthusiasts are ticked off and plastering MSI's forums with complaints.

Solution? Clockgen supports a "ghost" mode which basically allows you to use a config file to set the HTT, and you can run it at startup so once you get into Windows, your clock is bumped up to your chosen setting. So I sat down with a spreadsheet and started doing some benchmarks at various speeds. I tested a range of different overclocks from 1980Mhz to 2403Mhz. My 3DMark03 score went from 10492/958 to 10930/1113. And I was particularly impressed with my time to compile a HL2 map, it dropped from 10m3s to 8m10s.

So I will booting at a basic setting of 220HTT, 4x HT Link, 166 RAM Divider. And then using ClockGen in ghost mode at Win startup to boost my HTT to 267. Final settings: 267HTT x9 = 2403Mhz CPU, my RAM is at 218.5Mhz, 1T-2.5-3-3-6.

I'm going to start Prime95 right now ...
 
kronchev said:
And as many people will tell you (and I will confirm), a platform that cant prime95 for at least 24 hours straight is too unstable to be worth using.
Click to expand...
it's kind of a personal thing.. i find that if i can prime for just an hour without crashing, i will never encounter issues... hell, i used to run my rig <1min prime stable, and i only had lock-ups when i started gaming :p
 
You must log in or register to reply here.
Back
Top