View Full Version : did i oc too high?
I oced my amd 2000+ to about 1875(12.5x150)MHz... first i was happy because it booted up fine untill the windows xp load screen came up... i waited for several minutes but it just didn't want to start up windows. How come??? i have a MSI KT3 ULTRA 2 motherboard with a 256mb twinmos 2700 memory stick with 2-2-2-5-2 timings. plz help
I forgot to mention that my vcore: 1.85v; temp below 50*c
Grande Juan
09-18-03, 11:20 AM
crion try relaxing the memory timings a bit see if that helps also if you can and have'nt already lock the agp/pci. Are you using a optical usb mouse? These sometimes cause me problems when overclocking.
drshivas
09-18-03, 12:06 PM
Try overclocking in baby steps. Drop to stock, then up everything incrementally to your target. Test along the way. Check the stickies in the forums for guides on how to do this.
But it does sound like you oc'd too much. Something isn't liking the speed. It could the the CPU, memory or NB.
Good luck!
=[8c0re8]=
09-18-03, 02:20 PM
yes try an relax your memory timmings lower.
you should be able to get a more stable O/C with lower settings.I have that MB as my server/ftp an the highest i have had the fsb@
was 200 but that only lasted till i rebooted the system..After that
the HD got corupted an i had to do a clean install.
Average stable FSB for me was around180-185 both using a 2000+ an a 2200+ an Corsairs XMS pc3200 an Mushkins PC 3200.
note: when i hit 200 fsb it was with the mushkins an not the Corsairs.
Deathknight
09-18-03, 03:38 PM
One thing you may be running into is too high FSB. I think on the kt3 ultra that 1/5 divider kicks in around 160 or so fsb, so at 150 your pci speed is at 37.5. Thats not horrid but getting around danger territory for your HD.
From your symptoms it doesn't sound like that is the problem you are having, but as a word of warning I wouldn't mess around too much in the 150s, instead lower your mult and skip right into the 160s+ so you can take advantage of the 1/5 pci divider.
drshivas
09-18-03, 05:13 PM
I agree with Deathknight. Jump to 166FSB to get the 1/5 divider. Very important for stability and avoiding slow kill on your I/O devices and peripherals.
Amorphus
09-18-03, 08:36 PM
mind you , this is with a twinmos PC2700 stick of ram, so all you guys telling him to loosen up his timings... try again. :)
now, see, 1.85v is above most people's tolerances for voltage supplied to their procs. I myself don't like to go >1.8v
and your proc is reaching its limits, if its not booting into windows. slow it down. try 12*166, go for 1/5 and 2/5 dividers for PCI and AGP respectively.
Grande Juan
09-18-03, 09:47 PM
bows down to the overclocking god, Amorphus. ;)
Yeah lower FSB. Sounds like that is the prob.
Cyrix_2k
09-19-03, 10:27 AM
Uh, just curious, but what are your proc steppings?
Deathknight
09-19-03, 10:35 AM
Well 1.85v is not always over most people's tolerances, that depends on the default voltage of the chip and conisdering we don't know the stepping (heck it could be a tbred A for all we know) its not automatic to say that 1.85v is bad.
amok time
09-19-03, 10:40 AM
no smoke no fire, try lowering your fsb and raise your mult to say about 15 to start. If you go 15*133 you will be sitting good until you get better mem. Step up your FSB after that.
Have fun
thx alot for all answears... i will try these things out.
I have another question though! If i bought a stick of CORSAIR 256MB DDR XMS3000 would i be able to get the max out of them with my current mobo?
c627627
09-19-03, 05:21 PM
Originally posted by Deathknight
... its not automatic to say that 1.85v is bad.
I thought so too. Then
Originally posted by Gautam
Although the percentage increase in each is the same, voltage is a squared factor for heat dissipation. I'm positive that 2v is too much even for a DUT3C. 1.95v may very well be also. The risk doesn't justify the usage of voltages that high...
In other threads following this quote I believe Gautam argued 1.85 volts being the line for all CPUs regardless of default voltage.
No one ever reported long term degradation of highest stable OC when running a system below 1.85 volts on permanent bases.
They did report this when they were running CPUs over 1.85 volts.
Finally I got the time to try your ideas out. Another thing ran trough my mind... I forogot to mention that i have a palomino core. Here's some other facts that might not be so interesting of my CPU:
----------------------------------
Athlon XP Processor
Core Palomino
CPU Model 6
Manufacturing Process 180 nm
Approximate Transistor Count 37.5 million
Approximate Die Size 128 sq. mm
Performance Rating 2000+
Working frequency 1667 MHz
Package Type OPGA
Operating Voltage 1.75 V
Max Die Temperature 90° C
L1 Cache Size 128 KB
L2 Cache Size 256 KB
Multiplier 12.5x
FSB Frequency 266 MHz
Stepping Code AGOIA
Manufacture Year 2002
Manufacture Week 17
Production Batch F
----------------------------------
Back to the overclocking results... I didn't manage to boot properly with a fsb higher then 148. In windows i started 3dmark and watched it run trough some tests, and after about 4tests 3dmark just stoped :| after that i started SETI@home and after just a few seconds the computer reboots ...and i can't change my multipler :(
Should i buy better ram or the completepackage(mb,cpu and ram)? i fought of a abit nf7-s and a barton2500+ and some memory, right now i don't know for sure.
vBulletin® v3.8.7, Copyright ©2000-2012, vBulletin Solutions, Inc.