• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

Need help with First time A64 Overclock!

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.
osirus you might want to check everything you can for that kind of oc... if you were that unstable, i doubt windows would have loaded if your vcore was that off. check the lower voltages for 2.6... you might be able to get it stable at a lower end vcore instead of a higher. hopefully itl be lower so you wont have the sheer heat output that it would most likely produce
 
Here is my successfull attempt at 2.7

2.7.jpg
 
well I started prime95 when I left for work at 8am PST, and it failed at 11:59am PST
So I dont know how stable you would consider that.
 
There is no universal firm agreement how long prime95 has to run to be considered as "stable". A common consensus here is usually 24 hours.

But that is just a number, why not 12 hours, .... As environment such as room ambient temperature can also affect stability. The longer the time, the higher the confidence the CPU can be considered "stable" under changing environment conditions of voltage, ambient temperature, CPU/memory load, system background tasks, .... 12 hours would give certain level of confidence, and 18 hours would be more, and so on.

In your case, if it fails about 4 hour, why not lower the HTT from 300 MHz to, say 297 MHz, at which the CPU would run at 2673 MHz and see whether Prime95 can run longer, and so on, ..., to narrow down the search for max overclocking by moving the HTT slightly higher or lower.
 
thats a good point you bring up. This morning when I did the overclock, the room temp was in low teen's (c). (I know, hard to believe for cali). 2.6 is prime stable for 24hours, I stoped the test prior to trying to hit 2.7.
I'm VERY satisfied with my first AMD system, I would have to say I wont go back to intel anytime soon.
I would like to hit 2.8 eventally, but I dont think 1ghz overclock is possible with my current ram. Note - I'm still running at 3:2, I would like to get to 5:6 and be at 2.6, which Im sure my memory will play a part in that.
 
Even with slow memory, you may isolate the CPU and memory overclocking by pushing the CPU as high as possible (with enough cooling and safe voltage).

E.g. for testing CPU with max x9 multiplier at 2800 MHz

HTT = 311
LDT multiplier x3
so CPU would run at 2799 MHz

with memory_FSB_ratio 133 (i.e. 133:200 or 2:3)
CPU_divider = 14
memory would run at around 2799 / 14 = 200 MHz <- the PC3200 should do it

If not, then you may try
with memory_FSB_ratio 100 (i.e. 100:200 or 1:2)
CPU_divider = 18
memory would run at around 2799 / 18 = 155 MHz
 
well, it overclocks to about 265fsb on stock Vcore before it starts to become unstable in prime. So i up the Vcore some and it dosen't seem to help. i have tried up to 1.5v. i know my ram can go up to 260Mhz so of corse by this time its already on the 166 dividor. What else could be stopping the cpu from being stable?
HTT 3x
cpu ratio 9x
MSI Neo2
3000+
PQI 3200
any other info needed?
 
Last edited:
ok, first, set the AGP to 67.
then disable speedspectrum
dynamic overclocking - manual
dram config - 166:200 (if that doesnt work, try 133:200)
make sure you set it at 2T, and not 1T
fsb to 279
vcore to 1.55
vdimm to 2.85

with those settings
 
Osirus said:
ok, first, set the AGP to 67.
then disable speedspectrum
dynamic overclocking - manual
dram config - 166:200 (if that doesnt work, try 133:200)
make sure you set it at 2T, and not 1T
fsb to 279
vcore to 1.55
vdimm to 2.85

with those settings
my AGP is at 66
speedspectrum is disabled
dynamic overclocking is on manual
my dram is at 166:200
im at 2T (1T is unstable)
at 279 fsb with vcore at 1.55 it fails prime within one min
vdimm is at 2.6 (ram is stable at 2.6v up to 260fsb, tested ram before cpu)

those settings didnt seem to work it was worth a try though. One thing did catch my eye, my vcore is set in bios at 1.40v and on the monitoring software and CPU-Z shows that its at 1.36 to 1.38v stable though at 260fsb. when i upped the voltage to 1.55 it only showed it as 1.42 to 1.44v unstable at 265fsb.
 
Last edited:
ok, set the AGP to 67. i had problems at high FSB until i set it at 67.
have you tried 133:200?
set your vcore to 1.55 and vid % at 3.3%
vdimm at 2.85 (it wont hurt the ram)

give those a shot.
 
busy lately, umm i think i found the problem. i mentioned before how i could up the voltage but it seemed to do nothing well its that "vid %" option. Its a little hard to understand what it does. anyway upped it to 3.3% and away the vcore went. Kinda odd having 2 options that both affect the Vcore.

thanks for the help Osirus.
 
Well I have a memory question now...
Im going to be upgrading, or downgrading.. depending on who you talk to.
Im moving from 1gb of Geil Golden Dragon PC3200 Ram, to 512mb (2 x 256) Geil Golden Dragon PC3700 Ram... with timings of 2.5-7-4-4

I want to know what my odds are of getting a stable 2.7ghz are, or even getting more memory effecincy with my 2.6ghz setup...

Mind you that I am running 133:200 2T right now. Im hoping that I will be able to run 166:200 2T @ 2.6ghz

Also, am I going to notice a differnce if I wanted to run 1:1? and what speed or FSB would my memory have to be to use 1:1?

Thanks!
 
Back