Re: ASUS Crosshair help

John wrote:
I'm looking to overclock my CPU I was wondering on some help with the settings. I'm not used to so many. I have a stock cooler but I have 6 fans keeping it cool

I have a x2 6000+ cpu
4 gigs of corsair ram 5.5.5.12 1.9v

There is not a lot of headroom left, on a 90nm 3GHz AMD processor.
You can change your HT bus multiplier, from 200x5, to 200x4, then
start cranking the CPU clock from 200 to something higher. You
probably won't need to change any RAM timings at all, as the wheels
will likely fall off your processor, before you get too far. In
terms of cranking the 200MHz CPU input clock, try it 5MHz at a time.

To experiment in Windows, you are in luck. The Nforce 590 is supported
by Clockgen.

http://www.cpuid.com/clockgen.php

What you do, is set up the PLL and tell it you have an Nforce 590. The
dial box should then be available. (Note - I'm only going on how I set
up my ICS 952607 here, and I don't have an Nforce board, and the details
may be slightly different. Maybe the detection will be a bit more
automatic for you.) This is the dial box.

http://www.cpuid.com/pics/cgpllcontrol.png

The upper slider should control the CPU clock. Note that you cannot
use "clockgen" in Windows, until you've entered the BIOS and turned
down your HyperTransport multiplier to x4 (so HT runs at 800MHz, and
you have room to increase the CPU clock to 250, before 250x4 reaches
the 1000 limit).

With the HT multiplier turned down, and back in Windows, you can increase
the CPU clock 5MHz at a time. Click "apply selection" when you're happy
with the slider value. The way that clockgen works, is it increases the
clock 1MHz at a time, waiting a short interval between tries. You'll see
the "busy cursor" while it is doing this. If you are running CPUZ, you'll
notice the clock being ramped while clockgen is busy. For example, for
a large step change in frequency, clockgen will need about 30 seconds
to do it. The author of clockgen does it that way, on the assumption
that the PLL can only be "pulled a little bit at a time", without
crashing the CPU.

OK, now you have basic clock control. The second tool you'll need is
Orthos. Orthos will run two copies of the Prime95 Torture Test.
Orthos will stop and show an error message in the log window, at
the first sign of trouble. Start Orthos running, until you see
the paging has stopped and the test has started. Look for progress
messages in the log window.

http://sp2004.fre3.com/beta/beta2.htm

Let it run for a minute, so it has done a bit of testing. If it
stops immediately, your computer is not stable enough to do
anything with. You need to work on your voltages and timings etc.
in that case. You cannot be an overclocker, unless you know how
to set up a computer. It won't work safely or effectively
otherwise (what good is a computer that always crashes ?).

Let's assume it has passed the test for at least a minute. Now, go to
Clockgen, and increase the CPU from 200MHz (nominal) to 205MHz.
Click Apply Selection. Wait another minute or two. Did Orthos
stop and report an error ? If so, you've overclocked too far,
for the current Vcore selection and other conditions.

This will give you a quick taste of overclocking, without having
to do a lot of work. You would be advised to also watch the CPU
temperature, as it'll get hot during this test. The 6000+ is
rated for 125W power dissipation, and as you overclock, it will
draw more than that. Hot times await...

The purpose of changing the CPU clock only a little bit at
a time, is so you won't crash the computer outright. When Orthos
errors out, that is how you know you're getting near "the edge".
To go past "the edge", at that point you'll need to go back to the
BIOS and fiddle with things like Vcore. The purpose of Clockgen,
is to avoid having to drop down to the BIOS, and reboot so many
times. But remember that increasing Vcore, pushes up the power
dissipation, so sooner or later, your CPU will be an inferno.
And if pushed far enough, the Vcore MOSFETs around the CPU,
will also start to get hot.

Paul
.

Relevant Pages

  • Re: About Power Manager
    ... Depends on the capability of the cpu. ... Power states that affect other devices would be managed by the higher level ... If I move the CPU into idle mode, I will reduce the CPU clock for power s ... Is it OK for reducing CPU clock in the OEMIdle? ...
    (microsoft.public.windowsce.platbuilder)
  • Re: Stolen and degraded time and schedulers
    ... I presume you mean monotonic real time.) ... but I don't think that's terribly useful. ... work_time is probably most naturally measured in cpu clock cycles rather ... In the same way that clock modulation gates the cpu clock, ...
    (Linux-Kernel)
  • Re: Overclocking the c64
    ... >> of glue logic just to make CPU run at 2MHz and all the rest at ... > The reason that the CPU clock needs to be separated is that on a stock ... > more than just a little bit you would loose video and probably also ... the crystal determines both the CPU clock and video timing ...
    (comp.sys.cbm)
  • Re: timing a method
    ... That's stated to be a count of CPU clock cycles, ... >>be independent of CPU clock rate and drift, ... and Dr John, this is all said and good. ... So no reasonably accurate is going to be reasonably ...
    (alt.comp.lang.borland-delphi)
  • In the midst of chaos, an adventure can be found Topic Summary: A beginners journey into the realm o
    ... I've built it up and am installing windows. ... AMD CPU Driver & Optimizer. ... Quick restart then the mobo drivers, another restart and on go the ... I install ASUS PC Probe and the AI Booster utilities. ...
    (comp.programming)