View Full Version : How to reach your max stable fsb quickly, without hosing your OS.
RichBa5tard
10-16-2003, 03:02 AM
It seems to me like many of you people still use Prime95 / 3Dmark or any other windows based app to test stability while searching for the highest stable fsb. I'm not saying this is wrong, but there is a far faster way to find your highest stable overclock, without the chance of hosing your OS.
Step 1:
download (http://www.memtest86.com/#download0) memtest86 floppy or CD ISO image, and write the floppy/cd. Make sure "boot from cd/floppy" has a higher priority in your bios than your HD.
Step 2:
Overclock. :) Increase FSB by a few hz, change memtimings, whatever floats your boat.
Step 3:
Loading memtest86 only takes a few seconds. When it's loaded, select to loop test 5 (press "c","2","5","5","0","esc" i think). If it passes test 5 two or three times without generating errors, you've got a very, very good chance your system is stable. One loop should be a minute, depending on your ram size/speed. It only checks for memory errors, but experience tells if your cpu/chipset can't cope the high fsb, it will generate errors too.
Step 4:
As long as no errors where dedected, reboot and go back to step 2.
Step 5:
If you think you've hit the ceiling, boot from OS and start stressing with prime95 / 3D apps to ensure stability.
I've been using this method for years. It may not be the best method if our aiming for the highest 3Dmark/overclock (surely, your system won't crash instantly when memtest generates only a handfull errors, those can be corrected), but it's an excellent method to quickly find your optimal 24/7 overclock. If you got multiple motherboards to test on a limited time, this is by far the best method imho. :)
I'm sure many of you guys already know this, but I haven't seen it posted anywhere. I've stumbled upon it by accident and found out test 5 was the hardest to pass, so I just want to share my knowledge.
HEy,very nice. Why not make it sticky/pin it?:P
Soulburner
10-16-2003, 06:44 AM
To me, stable is when it can run 3DMark even if it fails memtest and Prime....:D
Neptune
10-16-2003, 06:54 AM
But then you got a fair chance your system willl crash later on.
Soulburner
10-16-2003, 06:58 AM
Originally posted by Neptune
But then you got a fair chance your system willl crash later on.
Yeah but that doesn't matter, I only run it at max for a score then it goes back to daily settings.
Originally posted by Soulburner
Yeah but that doesn't matter, I only run it at max for a score then it goes back to daily settings.
i agree
RichBa5tard
10-16-2003, 07:16 AM
Originally posted by Soulburner
Yeah but that doesn't matter, I only run it at max for a score then it goes back to daily settings.
And what's your daily setting? Stock speeds? If not, this is a fast way to find a stable daily setting. ;)
Soulburner
10-16-2003, 07:24 AM
Just 250 5:4 2-2-2-5 for 24/7. :D
Gaming doesn't really need much more than that.
macci
10-16-2003, 07:53 AM
I've got Win98 and Winxp on same drive so I usually just load Win98 safe mode and test how far stuff will go (using setfsb/pifast/superpi combo) and then load WinXP/2000 at safe speeds :)
..better go and reinstall Win2000 :D
To me running just to get a benchmark and then downlclocking to a stable speed doesnt seem right. Its cheating. The idea is to get a high stable clock and produce a benchamrk, and have it stable. Thats like running a processor at 4.9GHz just to say you got there, even though it wasnt stable :(
RichBa5tard, Nice thread :D
Craig
Ansur
10-16-2003, 08:00 AM
Thanks for sharing this with us!
I'll get right on using this handy piece of software to determine my max fsb ;)
saaya
10-16-2003, 08:02 AM
i dont trust memtest at all... says my system is fine but at the same speed prime crashes after 4 minutes and after a few days up and running i get a bsod.
bowman1964
10-16-2003, 08:03 AM
Well that how i do it to.
it is a very effecient way to reach your highest mem fsb.
even on a somewhat stable system i have seen memtest find errors on the mem when you test.normally if i pass 3 loops of test 5 it never gives me any trouble.most of the complates from guys getting kicked back to the desktop running 3d benches is from memory not stable and mem test will normally find it.
good thread:D
faruquehabib
10-16-2003, 08:04 AM
Originally posted by CCW
To me running just to get a benchmark and then downlclocking to a stable speed doesnt seem right. Its cheating. The idea is to get a high stable clock and produce a benchamrk, and have it stable. Thats like running a processor at 4.9GHz just to say you got there, even though it wasnt stable :(
RichBa5tard, Nice thread :D
Craig
i second this
Soulburner
10-16-2003, 08:40 AM
Originally posted by CCW
To me running just to get a benchmark and then downlclocking to a stable speed doesnt seem right. Its cheating. The idea is to get a high stable clock and produce a benchamrk, and have it stable. Thats like running a processor at 4.9GHz just to say you got there, even though it wasnt stable :(
RichBa5tard, Nice thread :D
Craig
No, it is not cheating. My scores in sig are 24/7 stable.
The point of the benchmark is to show what your system can do. It doesn't mean at all that you have to run those settings every day.
MrIcee
10-16-2003, 08:48 AM
I've found a solution to prevent hosing your OS or corrupting bootup files, etc....when pushing high FSB and using tight timings, particularly on AMD boards....is to set your hard drive(s) to PIO mode from DMA. 9 out of 10 times the hosing can be attributed to the hard drives direct memory access during the high FSB/tight timings trials and tribulations.:)
This has saved my OS's for quite some time now when going "over the top" in testing max FSB.
Randi:D
bowman1964
10-16-2003, 08:48 AM
Originally posted by Soulburner
No, it is not cheating. My scores in sig are 24/7 stable.
The point of the benchmark is to show what your system can do. It doesn't mean at all that you have to run those settings every day.
i dont think he was refering to you personally.
it is some like to brag they are running say 4.9g and really it barely will even post at those speeds for say.thats all.
there has always been 2 sides to that fence,and i see both sides.
Holst
10-16-2003, 10:07 AM
I just take smallish steps (5-10fsb) untill superPI wont run 1M then I drop back a couple of FSB.
I havent lost an OS for a whyle, and I keep a ghost anyway (that reminds me I need to backup)
As for stock/benching speeds.
Most of the time I run a little below my max clocks, usually 5fsb and 100mhz this allows lower Vcore (keeps my CPU alive) and pretty much guarantees stability. obviously if I crash at those speeds ill go back further..
I dont use memtest as I dont have a floppy on my PC.
Sarcastro
10-16-2003, 12:08 PM
Holst, you can download a bootable CD aswell. I use that one because it is faster than a floppy drive.
RB, I've been using your method eversince you first told me about it, it's great!:)
JCviggen
10-16-2003, 01:46 PM
I very rarely hose an OS even though i do crash them on a regular basis lol... memtest screw that that is a waste of time and it only works if you make small FSB jumps...and even then theres a chance it doesnt help you and you still hose it on the next jump *cough* VIA *cough*.
The thing is, I always have more than one OS,,, i just ZIP the windows folder. If you hose it... load the other windows, overwrite the f*cked one and there you go
Marci
10-16-2003, 03:49 PM
Or use a Globe Micro hd2 (AKA Pulseman), about £40... mini-PCI card which backs up CMOS and drive FAT or equivalent thereof, and provides hardware data recovery on the fly every boot up if required...
So far it's saved me from 100% of near fatal OS screwups over the past 2 weeks from pushing the FSB too hard... and even resurrected a dual install of Win2k & WinXP on the same drive... (trashed the 2k first, rebooted and didn' t pay attention and it booted into XP and trashed that too!)
http://www.responsecomputing.co.uk/pulseman.htm
I never destroyed any of my win2k installations :confused: (I know winxp is dying like a champ...)
I wonder how it's done ;)
I am not a missy when testing max fsb, but even in worst case it booted fine after decreasing unstable settings a little...
I only get errors while windows load due to cpu/bios d10/ and lockups through too ambitious voltages on cpu/vdd ;)
P.S. I always do use my max settings 24/7... it's equal if it is maxed out or just 2% under that limit :rolleyes: cause you change hardware too fast to see ya cpu/ram/vidcard dying :D
Saffire
10-16-2003, 05:55 PM
S'what I do, except the problem is figuring out what's the bottleneck is ;) Testing some new RAM at 220 as we speak, seems to be pretty stable...