"Doug" wrote
Can anyone tell me where the CPU disconnect function is on the NF7-S. I
think thats whats causing my instability @ 200mhz.


Hi,
isn't it on the softmenu page, along with FSB settings etc?, I have it
enabled (default) and have had no *instability* to speak of (Prime95 errors
but no crash). seems to be running well on 1.75v though. . .

Wayne][

I disabled the disconnect function last night and was able to up my fsb by
7mhz :) now 218x11 on a barton.

--
Les Ross
Certified by a
Professional
"Wayne Youngman" wrote in message
...

"Doug" wrote
Can anyone tell me where the CPU disconnect function is on the NF7-S. I
think thats whats causing my instability @ 200mhz.


Hi,
isn't it on the softmenu page, along with FSB settings etc?, I have it
enabled (default) and have had no *instability* to speak of (Prime95
errors
but no crash). seems to be running well on 1.75v though. . .

Wayne][

"Wayne Youngman" wrote
I just got to the stage where I am 100% satisfied that my system is
working
great at stock settings. I have tested the memory timings, and I have
tested my FSB above 200MHz.

Now I am working on overclocking my cpu. When I first switched up to 11 x
200MHz from default 2500+ all I did to test the overclock was run MemTest
x86 from the boot floppy. Of course to really test the CPU overclock I
needed to run tests from WinXP itself, but I was a little anxious about
causing my boot-drive to become garbled from a dodgy CPU overclock (it's
happened to me before).

Anyhow even though it was working fine in MemTest x86 @ default Voltages,
once I got into WinXP (phew it booted fine) and ran Prime96 torture test,
it
would fail within a minute. . .pooh!

So it was time to up the Vco

1.65v @ 2.2GHz - Fail Prime95
1.675v @ 2.2GHz - Fail Prime95
1.7v @ 2.2GHz - Fail Prime95
1.725v @ 2.2GHz - Fail Prime95
1.75v @ 2.2GHz - Been running Prime95 for 10 hours so far!

Temps are up up course, hitting 46c Load so far, I had to crank up my
CoolerMaster Aero 7 LITE Fan to 100% to keep the temps down a few degrees.

I will leave prime95 running for a day (or two), then when I am happy it
is
Rock-Solid I will tinker with a higher FSB to add to the CPU overclock
(thinking 10.5 x 216MHz-FSB).

Is 1.75v ok to be running this chip at, as I said temps are ok so far. . .

1.75 or 1.775 for that matter (see below) should be fine long-term but
nothing is guranteed as soon as you raise it even one notch. That said,
I've only had the one 'overvolting failure' and that is still debatable
because of the circumstances.


I hope it passes the Prime95 test as I don't really wanna raise Vcore
above
1.75 if I can. . .

When you find the 0.025 that first runs Prime95 for a while, go straight
back out to BIOS and go up another 0.025 - it's almost bound to fail Prime95
after a few hours if you only have that 0.025V in-hand above 'known bad'.
0.025 will make FA difference to temps - you need to back off or rise 0.1 to
see any difference and even then it shouldn't be a huge amount with an
efficient cooling sytem.


Wayne ][

P.S I have CPU interface enabled. Is there some options that make a CPU
overclock less stable? (spread-spectrum etc)

Yes:
a) CPU interface will give you some slight advantage but I think it stops
the overclock too low to be worthwhile.
b) Kill off both 'spread spectrums', all they do is stop RF interference at
the expense of your stability. If you get radio interfertence then improve
your shielding, spread spectrum sucks..!
c) Kill off the CPU disconnect function - another stability killer.
d) Kill off 'fast writes' there's barely any gain from it and it's a
stability killer.

As always, YMMV, suck it and see for yourself with a bunch of benchmarks
that emulate your desired usage pattern - but these are a good starting
point.

[UK]_Nick...

"Wayne Youngman" wrote in message
...

"Doug" wrote
Can anyone tell me where the CPU disconnect function is on the NF7-S. I
think thats whats causing my instability @ 200mhz.


Hi,
isn't it on the softmenu page, along with FSB settings etc?, I have it
enabled (default) and have had no *instability* to speak of (Prime95
errors
but no crash). seems to be running well on 1.75v though. . .


Nah, it's on the memory settings page along with 'spread spectrum' Etc, a
few down from the spread spectrum pair, maybe two up from the bottom I
think.

Ciao...

[UK]_Nick...

"Wayne Youngman" wrote
isn't it on the softmenu page, along with FSB settings etc?, I have it
enabled (default) and have had no *instability* to speak of (Prime95
errors
but no crash). seems to be running well on 1.75v though. . .

Hint - CPU FSB overclock it says on the righthand side of the page when you
enable "CPU Interface" - disable the sod until you're 100% stable, then
maybe try it whilst you're using the machine normally - if things go sour,
switch it off again & see if things get better. If it seems to give you
mucho extra performance (it doesn't IMO) then it may be worth backing off
the overclock to get it stable. Some tweaks are like that but not very
many.

& BTW, "CPU INTERFACE" _isn't_ "CPU DISCONNECT" that's on the memory page...
LOL

;-)

[UK]_Nick...

Double BTW: Why did you stop at 2.2GHz..? Most of those CPUs will do more
than that, weird if it won't and I'd suspect something else before the CPU.
My new Barton 2500+ (AQXEA) seems stable at 12x200MHz for 2400MHz but not a
midges more on the FSB - which suggests to me it'll be on the ragged edge at
that so I may back it off to 12.5x192 like the last one.

Could also be this particular NF7-s v2.0 and I'll be swapping the Barton out
for a Tbred B known good at 220MHz FSB to crosscheck that. Starnge that my
highest FSB clocking NF7 was a v1.2... Was being the operative word. :-(
But ya can't win 'em all and that's the very first mobo I lost to
overclocking...

Ciao...

[UK]_Nick...
--
Nick M V Salmon Master Mariner MN(Retd.)
Mailto: My four
http://www.nmvs.btinternet.co.uk/index.htm

"Nick M V Salmon" wrote
Double BTW: Why did you stop at 2.2GHz..? Most of those CPUs will do
more
than that, weird if it won't and I'd suspect something else before the
CPU.
My new Barton 2500+ (AQXEA) seems stable at 12x200MHz for 2400MHz but not
a
midges more on the FSB - which suggests to me it'll be on the ragged edge
at
that so I may back it off to 12.5x192 like the last one.

Could also be this particular NF7-s v2.0 and I'll be swapping the Barton
out
for a Tbred B known good at 220MHz FSB to crosscheck that. Starnge that
my
highest FSB clocking NF7 was a v1.2... Was being the operative word. :-(
But ya can't win 'em all and that's the very first mobo I lost to
overclocking...


Hi Nick and all persons here present,

Thanks for all the helpful feedback, quite a few interesting things there
(spread spectrum etc). The reason I *paused* on 2.2GHz is that it takes me
10 times as long to be confident that it is working smoothly. 2.4GHz sounds
amazing, lol I can really notice the difference even between 2500+ and 3200+
speeds (remember I been using a PIII 550MHz). What *PR* rating would a
2.4GHz be? 3500+ ?. Ok so you saying that I should try 11.5x200 and 12x200,
hehe you devil, I was just feeling like I got to the top of the hill, and
now I see that there is still more to climb!

Hmm I am running 1.75v now (Prime95 pass 48hours) I wonder if I could make
2.3/2.4GHz using that Vcore? I doubt it, but of course I will have to try.
Nick didn't you say that it was safe up to 1.775 Vcore?

Wayne ][

"Nick M V Salmon" wrote
Hint - CPU FSB overclock it says on the righthand side of the page when
you
enable "CPU Interface" - disable the sod until you're 100% stable, then
maybe try it whilst you're using the machine normally - if things go sour,
switch it off again & see if things get better. If it seems to give you
mucho extra performance (it doesn't IMO) then it may be worth backing off
the overclock to get it stable. Some tweaks are like that but not very
many.


Hi,

The only reason I had it enabled was it gave me quite a *boost* in Sandra
Memory tests. I'm not sure what you mean by *instabilities*. The only way
I know something isn't working is from MemTest errors and Prime95 fails. Do
you mean like sudden freezes and stuff?.

So what exactly will *disabling* CPU Interface give me? the chance for a
higher FSB, CPU overclock, memory timings? eh :P

Wayne ][

"Nick M V Salmon" wrote
Nah, it's on the memory settings page along with 'spread spectrum' Etc, a
few down from the spread spectrum pair, maybe two up from the bottom I
think.



Yup my silly! oops :P

Wayne ][

"Wayne Youngman" wrote
Hi Nick and all persons here present,

Thanks for all the helpful feedback, quite a few interesting things there
(spread spectrum etc). The reason I *paused* on 2.2GHz is that it takes
me
10 times as long to be confident that it is working smoothly. 2.4GHz
sounds
amazing, lol I can really notice the difference even between 2500+ and
3200+
speeds (remember I been using a PIII 550MHz). What *PR* rating would a
2.4GHz be? 3500+ ?.

Thereabouts although it does depend what FSB you're running. If you look at
the high-end AMD PRs mathematically then you'll know that they don't even
come close to linear progression with MHz, even within the same FSB group
like the 333FSB Barton 2500+, 2800+ and 3000+ - this is partly because only
the multiplier changes, so the FSB gets proportionally lower the faster the
CPU runs. IMO the 3000+ is badly mislabelled, I think they needed another
0.5 on the multiplier to give it a reasonable performance rating...


Ok so you saying that I should try 11.5x200 and 12x200,

Try 11.5x200 & run Prime95 for a short while, if it runs for ten minutes
then try 12x200 and run Prime95 again. I have one Tbred B that will do
11.5x200 for 2300MHz at 1.6Vcore, letalone 1.75...


hehe you devil, I was just feeling like I got to the top of the hill, and
now I see that there is still more to climb!

Best way to do it is push it until you get errors quite quickly at the
maximum Vcore you're willing to run, then back off on the overclock some way
so that Prime95 runs stable longer term. You're doing it a bit backwards
IMO.

NB: I'd advise you to keep written notes of what works because there are so
many variables it's sometimes hard to get back to a nice stable setup after
a CMOS reset.

At some point in the testing you usually find that it doesn't much matter
how much extra Voltage you shove through it, it won't do any more speed
without throwing errors - that's the true 'ragged edge' maximum but I never
run at that, they're never long-term stable with the wide variances in
temperatures we get from day-to-day. I usualy back off on the core Voltage
some way


Hmm I am running 1.75v now (Prime95 pass 48hours)

That's about as stable as it gets..!


I wonder if I could make 2.3/2.4GHz using that Vcore?
I doubt it, but of course I will have to try.

Try it, you'll never know until you do...


Nick didn't you say that it was safe up to 1.775 Vcore?

Yes, IMO. AMD sold some 0.13 micron Tbred B CPUs preset at 1.65Vcore. (my
2400+ Tbred B defaults to 1.65Vcore) That is a figure at which the CPU
should be good for circa ten years continuous use. Since the spec' allows
some latitude in that, I suspect you'd still get circa ten years out of a
CPU at 10% overvolt which is 1.76V for your CPU but who in hell reading this
NG is likely to be using the same CPU in ten years..?

Soooo, more than 1.76V should be good for plenty long enough for most of
us - how much more you ask..? How are your cojones & how much would you
cry if it went phut and you had to replace it is the only valid reply..!
I'm currently running my 2400+ 24x365 at 'only' 1.85Vcore simply because it
gets too damn hot if I use any higher Vcore, I shall doubtless shove it back
up to 1.95Vcore at which it will do 2.3Ghz stable when the weather cools
right off - I sure wouldn't cry if it went phut 'cause I can replace it with
a _better_ CPU for well under £50...

Ciao...

[UK]_Nick...