Sign in with
Sign up | Sign in
Your question

Unstable Athlon 2800+ with no overclocking, WHAT?

Last response: in Overclocking
Share
April 9, 2004 4:24:28 AM

Archived from groups: alt.comp.hardware.overclocking (More info?)

hi
just got my new system...
its not too great but it was cheap...

its

Athlon XP 2800+ Thoroubread Rev B
Jetway N2PAP400 Ultra motherboard
ATI Radeon 9600 Pro graphics card
512mb ADATA Ram, DDR400
2 case fans, one in at the front bottom 1 out at the back top
Dual fat QTEC 550 Watt PSU

Mem timings 11 3 3 2, tried these are lowest i could get... was not stable
on high settings either

My question really is what if anything is going on with my overclocking? and
why is it so unstable even at stock speeds?

At stock speeds programs like Distributed.net CPUBURN and Prime95 will crash
or freeze up my computer... 3dmark/seti@home will loop for 24 hours
straight without a hitch... but other things are so unstable its dis
heartening...

ive tried overvoltaging memory - no gain so put it to standard
overvoltaging vcore to 1.75v (default 1.65) - went higher but was getting
hot (stock cooling...) currently running it on stock cooling which is a
heatsink rated to 3200+... running 50 idle and a nasty 66 highest ive seen
under load before a freeze...

increasing multiplier... and overclocking some... made it rather more
unstable despite varying vcore alterations

underclocking, tried underclocking... got machine to be more stable and slow
(took it down to 1.8ghz)

decreasing multiplier and increasing fsb... no less stable but better
performance...

my question is why is it so damn unstable???

ive seen things that say i should be able to get my vcore to 1.9 volts... is
it me or is this insane and likely to destroy my processor? or would this be
ok for better cooled chips??

when increasing the vcore on a processor how should you increase it and when
do you know when to stop increasing?

thanks

alan
Anonymous
a b K Overclocking
April 9, 2004 5:47:01 AM

Archived from groups: alt.comp.hardware.overclocking (More info?)

Does your BIOS have a "Default" setting that sets everything to the stock or
"safe" values? If so, use that. If not, use the "Default" settings from
the manual. Then check your heatsink installation; the actual die
temperature could be much higher than the reading you get from an external
sensor. Then work from there. If even thoses settings don't work, try using
another CPU, switching the memory, try a different power supply, etc. If
this doesn't help, try a Pentium 4.

--
Phil Weldon, pweldonatmindjumpdotcom
For communication,
replace "at" with the 'at sign'
replace "mindjump" with "mindspring."
replace "dot" with "."

"Alan" <agoodm@removebtopenworld.com> wrote in message
news:c54qfs$m7j$1@titan.btinternet.com...
> hi
> just got my new system...
> its not too great but it was cheap...
>
> its
>
> Athlon XP 2800+ Thoroubread Rev B
> Jetway N2PAP400 Ultra motherboard
> ATI Radeon 9600 Pro graphics card
> 512mb ADATA Ram, DDR400
> 2 case fans, one in at the front bottom 1 out at the back top
> Dual fat QTEC 550 Watt PSU
>
> Mem timings 11 3 3 2, tried these are lowest i could get... was not stable
> on high settings either
>
> My question really is what if anything is going on with my overclocking?
and
> why is it so unstable even at stock speeds?
>
> At stock speeds programs like Distributed.net CPUBURN and Prime95 will
crash
> or freeze up my computer... 3dmark/seti@home will loop for 24 hours
> straight without a hitch... but other things are so unstable its dis
> heartening...
>
> ive tried overvoltaging memory - no gain so put it to standard
> overvoltaging vcore to 1.75v (default 1.65) - went higher but was getting
> hot (stock cooling...) currently running it on stock cooling which is a
> heatsink rated to 3200+... running 50 idle and a nasty 66 highest ive seen
> under load before a freeze...
>
> increasing multiplier... and overclocking some... made it rather more
> unstable despite varying vcore alterations
>
> underclocking, tried underclocking... got machine to be more stable and
slow
> (took it down to 1.8ghz)
>
> decreasing multiplier and increasing fsb... no less stable but better
> performance...
>
> my question is why is it so damn unstable???
>
> ive seen things that say i should be able to get my vcore to 1.9 volts...
is
> it me or is this insane and likely to destroy my processor? or would this
be
> ok for better cooled chips??
>
> when increasing the vcore on a processor how should you increase it and
when
> do you know when to stop increasing?
>
> thanks
>
> alan
>
>
Anonymous
a b K Overclocking
April 9, 2004 6:09:03 AM

Archived from groups: alt.comp.hardware.overclocking (More info?)

> There is no such CPU as an XP2800+ Tbred rev. B. Tbred's only went as far
as
> 2600+. It's either a Barton or you've been ripped off. How much L2 cache?
> Run CPU-Z or CPUID and see what it says. 256KB = Tbred, 512KB = Barton.


No i think your wrong there, there is a XP2800+ tbred rev B, not many were
produced, (http://www.hardwarezone.com/articles/view.php?cid=2&id=...)
Related resources
Anonymous
a b K Overclocking
April 9, 2004 1:47:31 PM

Archived from groups: alt.comp.hardware.overclocking (More info?)

"Alan" wrote
> Mem timings 11 3 3 2, tried these are lowest i could get... was not stable
> on high settings either
>
> My question really is what if anything is going on with my overclocking?
and
> why is it so unstable even at stock speeds?


Hi,
I just had a similar nightmare setting up my new P4 build, I couldn't get it
stable at stock speeds. In the end I worked out I had two *duff* sticks of
OCZ memory (or at least incompatible).

Did you run Memtest86 for 24 hours on *all* tests?
--
Wayne ][
Anonymous
a b K Overclocking
April 9, 2004 2:39:35 PM

Archived from groups: alt.comp.hardware.overclocking (More info?)

power supply, etc. If
> this doesn't help, try a Pentium 4.
>

Apart from the fact that the chip alone will cost as much as tthe whole AMD
system...

hamman
Anonymous
a b K Overclocking
April 10, 2004 1:20:44 AM

Archived from groups: alt.comp.hardware.overclocking (More info?)

"Alan" <agoodm@removebtopenworld.com> wrote in message
news:c54qfs$m7j$1@titan.btinternet.com...

[....]
> 512mb ADATA Ram, DDR400

[....]
> Mem timings 11 3 3 2, tried these are lowest i could get... was not stable
> on high settings either

You're lucky it will even boot at those settings - all my Adata memory is at
11-3-3-2.5 & seems to need 2.7V for stability at only 200MHz.


> My question really is what if anything is going on with my overclocking?
and
> why is it so unstable even at stock speeds?

Some other BIOS setting or piece of hardware.


> At stock speeds programs like Distributed.net CPUBURN and Prime95 will
crash
> or freeze up my computer... 3dmark/seti@home will loop for 24 hours
> straight without a hitch... but other things are so unstable its dis
> heartening...

Prime95 is a good stability test - if you can get it stable running Prime95
then it ought to be stable for most everything else.


> ive tried overvoltaging memory - no gain so put it to standard
> overvoltaging vcore to 1.75v (default 1.65) - went higher but was getting
> hot (stock cooling...) currently running it on stock cooling which is a
> heatsink rated to 3200+... running 50 idle and a nasty 66 highest ive seen
> under load before a freeze...

I wouldn't try pushing the Vcore very high with only stock cooling.


> increasing multiplier... and overclocking some... made it rather more
> unstable despite varying vcore alterations
>
> underclocking, tried underclocking... got machine to be more stable and
slow
> (took it down to 1.8ghz)
>
> decreasing multiplier and increasing fsb... no less stable but better
> performance...
>
> my question is why is it so damn unstable???

It sounds like something other than the CPU is causing instability - a
recent core Tbred B really ought to be absolutely rock stable at only
2.083GHz & 1.65Vcore.


> ive seen things that say i should be able to get my vcore to 1.9 volts...
is
> it me or is this insane and likely to destroy my processor? or would this
be
> ok for better cooled chips??

Many folks have run at 1.9Vcore for extended periods but with better cooling
tha stock - I have a number of systems at 1.85Vcore but almost all mine are
watercooled.


> when increasing the vcore on a processor how should you increase it and
when
> do you know when to stop increasing?

If the CPU is unstable then increase the Vcore a notch at a time until
either the CPU is stable or the temperature gets too hot, if it gets too hot
then back off the overclock instead - I won't run any CPU over 60C measured
by the motherboard - although that is well within AMD's spec' I've found
they get less stable above that - it does depend on the motherboard though
eg. Abit NF7s and KX7-333s read much lower than Epox 8K3As at the same speed
& Vcore.

Ciao...

[UK]_Nick...
--
Nick M V Salmon Master Mariner MN(Retd.)
Email: My four initials at dsl dot pipex dot com
http://www.nmvs.dsl.pipex.com/index.htm
April 10, 2004 2:06:29 AM

Archived from groups: alt.comp.hardware.overclocking (More info?)

"Nick M V Salmon" <spam@dump> wrote in message
news:407705d7$0$601$cc9e4d1f@news.dial.pipex.com...
> "Alan" <agoodm@removebtopenworld.com> wrote in message
> news:c54qfs$m7j$1@titan.btinternet.com...
>
> [....]
> > 512mb ADATA Ram, DDR400
>
> [....]
> > Mem timings 11 3 3 2, tried these are lowest i could get... was not
stable
> > on high settings either
>
> You're lucky it will even boot at those settings - all my Adata memory is
at
> 11-3-3-2.5 & seems to need 2.7V for stability at only 200MHz.

well actually i tell a lie here
one stick is ADATA and the other is some othger brand..,. but the modules on
the ram look identical and all the connections chips bumps and lumps are the
same...

> > At stock speeds programs like Distributed.net CPUBURN and Prime95 will
> crash
> > or freeze up my computer... 3dmark/seti@home will loop for 24 hours
> > straight without a hitch... but other things are so unstable its dis
> > heartening...
>
> Prime95 is a good stability test - if you can get it stable running
Prime95
> then it ought to be stable for most everything else.

ok, yes that is what i believed...
sometimes moving the mouse quickly while prime95 is running will cause an
error... is this normal?

> > ive tried overvoltaging memory - no gain so put it to standard
> > overvoltaging vcore to 1.75v (default 1.65) - went higher but was
getting
> > hot (stock cooling...) currently running it on stock cooling which is a
> > heatsink rated to 3200+... running 50 idle and a nasty 66 highest ive
seen
> > under load before a freeze...
>
> I wouldn't try pushing the Vcore very high with only stock cooling.

no this is my viewpoint too...

> > increasing multiplier... and overclocking some... made it rather more
> > unstable despite varying vcore alterations
> >
> > underclocking, tried underclocking... got machine to be more stable and
> slow
> > (took it down to 1.8ghz)
> >
> > decreasing multiplier and increasing fsb... no less stable but better
> > performance...
> >
> > my question is why is it so damn unstable???
>
> It sounds like something other than the CPU is causing instability - a
> recent core Tbred B really ought to be absolutely rock stable at only
> 2.083GHz & 1.65Vcore.

the stock for my tbred seems to be 2254mhz according to what the bios is
saying
i have the latest bios update flashed

> > ive seen things that say i should be able to get my vcore to 1.9
volts...
> is
> > it me or is this insane and likely to destroy my processor? or would
this
> be
> > ok for better cooled chips??
>
> Many folks have run at 1.9Vcore for extended periods but with better
cooling
> tha stock - I have a number of systems at 1.85Vcore but almost all mine
are
> watercooled.

ok, im considering getting a better air cooler... Volcano 12 to be
precise... these look to be very good... much much better than the stock

> > when increasing the vcore on a processor how should you increase it and
> when
> > do you know when to stop increasing?
>
> If the CPU is unstable then increase the Vcore a notch at a time until
> either the CPU is stable or the temperature gets too hot, if it gets too
hot
> then back off the overclock instead - I won't run any CPU over 60C
measured
> by the motherboard - although that is well within AMD's spec' I've found
> they get less stable above that - it does depend on the motherboard though
> eg. Abit NF7s and KX7-333s read much lower than Epox 8K3As at the same
speed
> & Vcore.

i do often see 60c after playing games on the current best settings... but
all seems to be very stable at 60c... o often get freezes at lower
temperatures then this

thanks for the information here, it is much appreciated

ill battle on!

alan
Anonymous
a b K Overclocking
April 10, 2004 2:47:01 AM

Archived from groups: alt.comp.hardware.overclocking (More info?)

"~misfit~" <misfit61nz@yahoomung.co.nz> wrote in message
news:Hiqdc.9688$d%6.173633@news.xtra.co.nz...
> Figurt wrote:
> >> There is no such CPU as an XP2800+ Tbred rev. B. Tbred's only went
> >> as far as 2600+. It's either a Barton or you've been ripped off. How
> >> much L2 cache? Run CPU-Z or CPUID and see what it says. 256KB =
> >> Tbred, 512KB = Barton.
> >
> >
> > No i think your wrong there, there is a XP2800+ tbred rev B, not many
> > were produced,
> > (http://www.hardwarezone.com/articles/view.php?cid=2&id=...)
>
> I stand corrected, thank you. However, as noted on that page "Critics have
> called the Athlon XP 2800+ announcement as only a "paper" launch since AMD
> has problems delivering the product into the marketplace". I have never
seen
> one advertised for sale at any of the websites I visit(ed) regularly and
> have never heard of one in the wild.
>
> If the OP actually has one then I suggest he pulls it and keeps it safe.
> It's probably one of only a few that ever got released. That could explain
> his problems with getting it to run properly.
>
> Cheers,
> --
> ~misfit~
>
>

I received one of these accidentally a few months ago - it seemed to run
fine for the day or two I had it, until I realised it didn't have the 512K
cache. The company I got it from sent me a new one, with the Barton core.
A rather unuseful contribution from me I'm afraid.
Rich.
Anonymous
a b K Overclocking
April 10, 2004 3:20:31 PM

Archived from groups: alt.comp.hardware.overclocking (More info?)

"Alan" <agoodm@removebtopenworld.com> wrote
> "Nick M V Salmon" <spam@dump> wrote
> > "Alan" <agoodm@removebtopenworld.com>
>
> well actually i tell a lie here one stick is ADATA and the other is some
> othger brand..,. but the modules on the ram look identical and all the
> connections chips bumps and lumps are the same...

Whatever - the Adata runs okay at 11-3-3-2.5, DDR Voltage 2.7V, try those
settings an see if it's anymore stable.


> > Prime95 is a good stability test - if you can get it stable running
> > Prime95 then it ought to be stable for most everything else.
>
> ok, yes that is what i believed...
> sometimes moving the mouse quickly while prime95 is running will cause an
> error... is this normal?

I've never seen that one myself - hmmmn, is that a USB mouse..? If so then
it may be the USB that's causing problems, but, it shouldn't be on an
NForce2 board as the PCI bus (& AGP port) are frequency locked - weird one.
:-?


> > It sounds like something other than the CPU is causing instability - a
> > recent core Tbred B really ought to be absolutely rock stable at only
> > 2.083GHz & 1.65Vcore.
>
> the stock for my tbred seems to be 2254mhz according to what the bios is
> saying
> i have the latest bios update flashed

Yeah, sorry, I was thinking Barton 2800+, not Tbred B 2800+ which does run
at 13.5x167 for 2250MHz.


> > Many folks have run at 1.9Vcore for extended periods but with better
> cooling
> > tha stock - I have a number of systems at 1.85Vcore but almost all mine
> are
> > watercooled.
>
> ok, im considering getting a better air cooler... Volcano 12 to be
> precise... these look to be very good... much much better than the stock

I've read good things about those on the web but I've never used one myself.


> > If the CPU is unstable then increase the Vcore a notch at a time until
> > either the CPU is stable or the temperature gets too hot, if it gets too
> hot
> > then back off the overclock instead - I won't run any CPU over 60C
> measured
> > by the motherboard - although that is well within AMD's spec' I've found
> > they get less stable above that - it does depend on the motherboard
though
> > eg. Abit NF7s and KX7-333s read much lower than Epox 8K3As at the same
> speed
> > & Vcore.
>
> i do often see 60c after playing games on the current best settings... but
> all seems to be very stable at 60c... o often get freezes at lower
> temperatures then this

Maybe software or driver problems then - if it was hardware then it would
freeze up more easily when hot...


> thanks for the information here, it is much appreciated
>
> ill battle on!

Only way to win. ;-)

Ciao...

[UK]_Nick...
--
Nick M V Salmon Master Mariner MN(Retd.)
Email: My four initials at dsl dot pipex dot com
http://www.nmvs.dsl.pipex.com/index.htm
!