MB Monitoring Utilities?

Per Beemer:
>My guess is that Easytune is trying to monitor one or more fans that are not
>actually installed as happened to me. Touching each fan's center with your
>finger will show which fan is which as you watch the Easytune fan page.
>Then you can lower the slider of the missing fan's so that it changes from
>red to blue. Of course you must have the fan control activated in the bios.
>My fans are only 3 pin so I had to set "voltage" as the fan control bios
>option. You can then select the advanced fan control in Easytune. If you
>cannot stop the noise you can always switch the sound off in Easytune.

That rings true to me.

I'll be pulling this thing apart in the next few days if/when I
go to Raid0 and I'll do it after rebuilding the sys.
--
PeteCresswell

I installed it and it works great for me, however, if I start changing some
of the settings, I assume I could lock up my PC easily.

For example, the FSB setting has a slider. If I cranked that up to maximum,
the PC could lock up?

--g

geoff wrote:
> I installed it and it works great for me, however, if I start changing some
> of the settings, I assume I could lock up my PC easily.
>
> For example, the FSB setting has a slider. If I cranked that up to maximum,
> the PC could lock up?
>
> --g
>

That is generally the idea. I was experimenting with my latest build
just yesterday, and that is what happened. Instead of Prime95 erroring
out, the computer just froze when I bumped up the FSB while in Windows.
So a freeze is possible. In my case, I was using SETFSB, because
there is no utility for my board.

Paul

When one changes one of those sliders, is the change made in the bios.

In other words, if the FSB is bumped up and the computer freezes, can I
simply reboot or do I have to put the setting back in the BIOS?

--g

geoff wrote:
> When one changes one of those sliders, is the change made in the bios.
>
> In other words, if the FSB is bumped up and the computer freezes, can I
> simply reboot or do I have to put the setting back in the BIOS?
>
> --g
>

I've never used any tool that changed the BIOS as a side effect.
But I understand it is possible to do it. I personally
wouldn't use a tool that has BIOS hooks, because that is
a little too much automation for my liking.

Paul

In general, in terms of percentage, say for the FSB, how much can the speed
be cranked up and still have a reliable system?

--g

geoff wrote:
> In general, in terms of percentage, say for the FSB, how much can the speed
> be cranked up and still have a reliable system?
>
> --g
>
>

I'm working on my system right now. The processor was selected,
to suit a motherboard with a low limit on FSB (1066).

2.6GHz/FSB800/2MB L2 - nominal is 1.35V (E4700)

With a BSEL mod and boosted Vcore (OFFSET pin mod)

3.46GHz/FSB1066 - 1.45V

Under those conditions, Prime errors out on one of the two cores,
after 15 minutes. Before the Vcore mod was added, I couldn't even
get the BIOS to start at that speed.

If I use CrystalCPUID, which can change the multiplier and the
voltage (but the voltage option is useless for this - VID changes
are clipped by the CPU internal logic), I can then run the following
condition

2.66GHZ/FSB1066 (multiplier 10X) - 1.45V

With those conditions, where the core is no longer overclocked,
but my FSB is, I've run Prime on the two cores for eight hours
without a problem. So that tells me the chipset, crappy as it is,
isn't to blame. I'm not FSB limited. I also get a small percentage
improvement in memory bandwidth (3000 to 3300MB/sec, depending on
settings and benchmark app). That is still quite a bit lower than
any regular motherboards.

So currently, that tells me I need a bit more Vcore, if I want
to pass Prime. As long as it hasn't hit a wall, and I don't have
enough data points to guess at that right now. (I cannot adjust
the clock in the BIOS, because the BIOS is known to misprogram
the clock generator. Even small overclocks fail. That is why
I used BSEL instead, as a canonical clock choice means the BIOS
doesn't have to change anything.)

I had hoped, when I started this project, I'd be able
to just set the clock to 266MHz, and magically I'd get
3.5GHz :-)

My last overclock (S478) went about as well. My P4 wouldn't overclock
very well either. The experience seems to be, once Intel fine
tunes the binning, the cheap processors lose their magic. When
a new generation of processor comes out, the first barrel-full
seems to be less limited. (Nehalem, get'em while they're hot.)

I wouldn't even have bothered overclocking in the first place
(because right now, everything is speedy enough at stock), but
I noticed when benchmarking that my memory bandwidth was terrible.
All I really wanted to do was bump the FSB and leave the core
alone, but this Asrock motherboard is fighting every inch of the
way. That is why I had to get out the soldering iron... To make
up for some moronic decisions by some BIOS designers. They gutted
EIST (disabled in BIOS and setting disappeared), and then prevented
multiplier changes, and the multiplier field says "change disabled
whenever EIST is enabled". Since they disabled EIST and removed it
from the BIOS, by their own logic, I should be able to change the
multiplier. Nope. So I have to change the multiplier in Windows,
and get the computer stable enough, so it can boot at 3.46GHz (only
realistic option with respect to the BIOS).

That is what I get, for wanting to keep my AGP card and save
a few bucks. I'm almost tempted to keep working on it, until
I break it.

Paul