KT7A v1.2 A9, Thorton 2000MHZ + 133FSB/15x multiplier ???

[Guest]

Archived from groups: alt.comp.periphs.mainboard.abit (More info?)

I'm using a Thorton 2400 which boots at 2000MHZ, and with the only (?)
useable setting of "above12X", I think, the FSB runs at 100MHZ, giving
an actual multiplier of 20x, verified by CPUID.

I'd like to use 15x with 133MHZ FSB to get better performance with
cetain applications.

I noticed some time (couple of years ?) ago jumpering diagrams for the
CPU socket to allow the necessary FSB multiplier.

Forgot to save these, or haver lost them !

If anybody can direct me to the correct info I'll be most grateful. I'm
quite satisfied still with the ABIT KT7A, having bios A9 u.g. done and
don't see the need to junk it yet.

Replies gratefully received,

Thanks in advance,

Tom.

(Not too interested in trying the 256K-512K cache mod, seems a bit
tricky, and instructions can't be found-or can they ? Overclockers ru
english version which I looked at has gone...)
--
Tom Goodman

Chairman
West London Wildlife Group
Regd. Charity No. 295834

Mob: 077 806 22 070
Fxd: 020 899 55 722
Skype IP Telephony: nosebag

 

[Guest]

Archived from groups: alt.comp.periphs.mainboard.abit (More info?)

On Fri, 09 Sep 2005 20:22:38 +0100, Tom Goodman wrote:

> I'm using a Thorton 2400 which boots at 2000MHZ, and with the only (?)
> useable setting of "above12X", I think, the FSB runs at 100MHZ, giving
> an actual multiplier of 20x, verified by CPUID.
>
> I'd like to use 15x with 133MHZ FSB to get better performance with
> cetain applications.
>
> I noticed some time (couple of years ?) ago jumpering diagrams for the
> CPU socket to allow the necessary FSB multiplier.
> If anybody can direct me to the correct info I'll be most grateful. I'm
> quite satisfied still with the ABIT KT7A, having bios A9 u.g. done and
> don't see the need to junk it yet.
>
Check my server below. The mods are in the Multiplier Cross ref. section.
17x (9x in bios) will work with a 133MHz fsb without mods, but for 2266MHz
you will probably need to raise vcore to about 1.70v.

> (Not too interested in trying the 256K-512K cache mod, seems a bit
> tricky, and instructions can't be found-or can they ? Overclockers ru
> english version which I looked at has gone...)

They can be found. May be here. But I thought AMD starting locking that
too.

http://fab51.com/index-e.html

--
KT133 MB, CPU @2400MHz (24x100): SIS755 MB CPU @2330MHz (10x233)
Need good help? Provide all system info with question.
My server http://wesnewell.no-ip.com/cpu.php
Verizon server http://mysite.verizon.net/res0exft/cpu.htm

 

[Guest]

Archived from groups: alt.comp.periphs.mainboard.abit (More info?)

Wes Newell wrote:
> On Fri, 09 Sep 2005 20:22:38 +0100, Tom Goodman wrote:
>
>
>>I'm using a Thorton 2400 which boots at 2000MHZ, and with the only (?)
>>useable setting of "above12X", I think, the FSB runs at 100MHZ, giving
>>an actual multiplier of 20x, verified by CPUID.
>>
>>I'd like to use 15x with 133MHZ FSB to get better performance with
>>cetain applications.
>>
>>I noticed some time (couple of years ?) ago jumpering diagrams for the
>>CPU socket to allow the necessary FSB multiplier.
>>If anybody can direct me to the correct info I'll be most grateful. I'm
>>quite satisfied still with the ABIT KT7A, having bios A9 u.g. done and
>>don't see the need to junk it yet.
>>
>
> Check my server below. The mods are in the Multiplier Cross ref. section.
> 17x (9x in bios) will work with a 133MHz fsb without mods, but for 2266MHz
> you will probably need to raise vcore to about 1.70v.
>
>
>>(Not too interested in trying the 256K-512K cache mod, seems a bit
>>tricky, and instructions can't be found-or can they ? Overclockers ru
>>english version which I looked at has gone...)
>
>
> They can be found. May be here. But I thought AMD starting locking that
> too.
>
> http://fab51.com/index-e.html
>

Wes,
You are a WIZARD !

Now running 133MHZ FSB x 9 in bios= actual 17x=2266.7MHZ; I put vcore up
to 1.70v in bios, and theother voltage setting from 3.2v to 3.6v just to
be safe (not sure what it does).

CPU-Z indicates 2266.7 MHZ, 17x multiplier of 133.3 MHZ, bus speed of
266.7 MHZ, and it detects vcore of 1.682 volts a bit less than I set in
bios, don't know why. MBM now shows 48C idling, and only 51C after
running CPU Bench 2003.
I got a blue screen and a memory dump after the first boot and attempt
to run a program, but this seems to be "normal" with XP Pro and NTFS,
subsequent boot ok as expected. No idea why this happens, but suspect
interaction with Zone Alarm/permissions.

Anyhow, all running smoothly with a painless >10% overclocking included
! MANY MANY THANKS.

I'm tempted to look into the 256-512 cache mod, as this ought to improve
performance still further (but by how much approx ?)

No I can implement the Nebula Digitv update to 3.5 beta which included
the ability to watch one program whilst recording (to MPEG2, the
broadcast compresion) another in the same multiplexed data stream. They
reckon a 2000MHZ Athlon XP would be the min spec, and I didn't think a
measly 100MHZ bus speed would be much good.

Sincere thanks,

Tom Goodman.

 

[Guest]

Archived from groups: alt.comp.periphs.mainboard.abit (More info?)

On Fri, 09 Sep 2005 23:44:04 +0100, Tom Goodman wrote:

> I'm tempted to look into the 256-512 cache mod, as this ought to improve
> performance still further (but by how much approx ?)

It all depends on the type app you run. Itf the app uses the extra cache,
it will be sped up. If it doesn't there will actually be a slowdown
because of the overhaed of the extra cache. Overall, I'd estimate about
2-3% increase.

--
KT133 MB, CPU @2400MHz (24x100): SIS755 MB CPU @2330MHz (10x233)
Need good help? Provide all system info with question.
My server http://wesnewell.no-ip.com/cpu.php
Verizon server http://mysite.verizon.net/res0exft/cpu.htm

 

[Guest]

Archived from groups: alt.comp.periphs.mainboard.abit (More info?)

Wes Newell wrote:
> On Fri, 09 Sep 2005 23:44:04 +0100, Tom Goodman wrote:
>
>
>>I'm tempted to look into the 256-512 cache mod, as this ought to improve
>>performance still further (but by how much approx ?)
>
>
> It all depends on the type app you run. Itf the app uses the extra cache,
> it will be sped up. If it doesn't there will actually be a slowdown
> because of the overhaed of the extra cache. Overall, I'd estimate about
> 2-3% increase.
>
Oh, well, in that case 2-3% hardly seems worth the risks and hassle
involved in pulling out the CPU, jumpering it, (still don't have the
connections) etc. The machine is running like a train now, and I've just
put in a 5 port USB 2 card to speed up peripheral work. I'll be keeping
the KT7A for some time now !

But, one wonders, if there is such a small gain, and possible slowup due
to overheads, why AMD bother with the 512 cache in the first place. It's
big chunk of silicon...

Tom.

 

[Guest]

Archived from groups: alt.comp.periphs.mainboard.abit (More info?)

On Sat, 10 Sep 2005 17:51:44 +0100, Tom Goodman wrote:

> But, one wonders, if there is such a small gain, and possible slowup due
> to overheads, why AMD bother with the 512 cache in the first place. It's
> big chunk of silicon...
>
It's an easy way to get just a little more out of the core. Adding the
extra cache is a simple matter compared to designing a new core, which
they aready had in the works with the K8.

--
KT133 MB, CPU @2400MHz (24x100): SIS755 MB CPU @2330MHz (10x233)
Need good help? Provide all system info with question.
My server http://wesnewell.no-ip.com/cpu.php
Verizon server http://mysite.verizon.net/res0exft/cpu.htm

 

[Guest]

Archived from groups: alt.comp.periphs.mainboard.abit (More info?)

Wes Newell wrote:
> On Sat, 10 Sep 2005 17:51:44 +0100, Tom Goodman wrote:
>
>
>>But, one wonders, if there is such a small gain, and possible slowup due
>>to overheads, why AMD bother with the 512 cache in the first place. It's
>> big chunk of silicon...
>>
>
> It's an easy way to get just a little more out of the core. Adding the
> extra cache is a simple matter compared to designing a new core, which
> they aready had in the works with the K8.
>
I see.
Still, I can't find the link for the wiring to try and get the 512 cache
anyhow, so I'll just stay the present improvement.

I have a further problem that I've not mentioned because I don't know
what the cause is and I'm reluctant to post in the wrong group. The
whole system "locks up" and neither keyboard or mouse respond. Reset is
necessary. If digital tv card is operative, then I notice CPU %
fluctuate and creep up to 100% in windows task manager, then the lock
up. sound sometimes "stutters" and the system locks up. (SB Live! card).
No overheating according to MBM, all drivers up-to-date and I suspect
hardware, but don't know how to start looking really. Sapphire Radeon
9200SE, DXdiag reports all ok. I've two 512MB memory modules and I
wonder of I should start by suspecting them ? Perhaps take one out at a
time and swap slots around. Memtest seems a recommended program, but I
see it can indicate faults which might not actually occur under "normal"
operating conditions. Replacing the whole 1GB of memory would cost about
£170, an expensive option. Where to start ? Software tests abound, but
hardware diagnostics seem thin on the ground.

tom.

 

[Guest]

Archived from groups: alt.comp.periphs.mainboard.abit (More info?)

In article <dg1toa$8gl$1@news.freedom2surf.net>, Tom Goodman
<tovmas@hotmail.com> writes
>
>I have a further problem that I've not mentioned because I don't know
>what the cause is and I'm reluctant to post in the wrong group. The
>whole system "locks up" and neither keyboard or mouse respond. Reset is
>necessary. If digital tv card is operative, then I notice CPU %
>fluctuate and creep up to 100% in windows task manager, then the lock
>up. sound sometimes "stutters" and the system locks up. (SB Live! card).
>No overheating according to MBM, all drivers up-to-date and I suspect
>hardware, but don't know how to start looking really. Sapphire Radeon
>9200SE, DXdiag reports all ok. I've two 512MB memory modules and I
>wonder of I should start by suspecting them ? Perhaps take one out at a
>time and swap slots around. Memtest seems a recommended program, but I
>see it can indicate faults which might not actually occur under "normal"
>operating conditions. Replacing the whole 1GB of memory would cost about
>£170, an expensive option. Where to start ? Software tests abound, but
>hardware diagnostics seem thin on the ground.
>

You do not say what operating system you are using. There are memory
considerations when the memory exceeds 512 MB in Windows ME and 98. See

http://support.microsoft.com/kb/253912/EN-US/
--
Nicholas David Richards -

"Où sont les neiges d'antan?"

 

[Guest]

Archived from groups: alt.comp.periphs.mainboard.abit (More info?)

Nicholas D Richards wrote:

>
> You do not say what operating system you are using. There are memory
> considerations when the memory exceeds 512 MB in Windows ME and 98. See
>
> http://support.microsoft.com/kb/253912/EN-US/

Sorry ! I should have said XP Pro SP2 updated to present.

 

[Guest]

Archived from groups: alt.comp.periphs.mainboard.abit (More info?)

In article <dg281q$cjg$1@news.freedom2surf.net>, Tom Goodman
<tovmas@hotmail.com> writes
>Nicholas D Richards wrote:
>
>>
>> You do not say what operating system you are using. There are memory
>> considerations when the memory exceeds 512 MB in Windows ME and 98. See
>>
>> http://support.microsoft.com/kb/253912/EN-US/
>
>Sorry ! I should have said XP Pro SP2 updated to present.

That is OK. Then the amount of memory should not be a problem.

Are you using a USB Mouse and/or keyboard?. There was/is a problem with
the USB through the Via chipset. The symptom was that USB devices would
freeze up although non-USB processing would continue normally.

There were many solutions suggested most of which did not work for me
until I installed a USB card that did not use a Via chipset. I am
posting this from that KT7A board, firmware A9 using an Athlon at 1200
MHz. The operating system is ME.

The other solution that worked was to disable ACPI, which involved a
registry hack and a lot of hassle to get it working. This is no use to
you as I suspect the registry hack is not available in XP.
--
Nicholas David Richards -

"Où sont les neiges d'antan?"

 

[Guest]

Archived from groups: alt.comp.periphs.mainboard.abit (More info?)

On Sun, 11 Sep 2005 19:41:12 +0100, Tom Goodman wrote:

> I have a further problem that I've not mentioned because I don't know
> what the cause is and I'm reluctant to post in the wrong group. The
> whole system "locks up" and neither keyboard or mouse respond. Reset is
> necessary. If digital tv card is operative, then I notice CPU %
> fluctuate and creep up to 100% in windows task manager, then the lock
> up. sound sometimes "stutters" and the system locks up. (SB Live! card).
> No overheating according to MBM, all drivers up-to-date and I suspect
> hardware, but don't know how to start looking really. Sapphire Radeon
> 9200SE, DXdiag reports all ok. I've two 512MB memory modules and I
> wonder of I should start by suspecting them ? Perhaps take one out at a
> time and swap slots around. Memtest seems a recommended program, but I
> see it can indicate faults which might not actually occur under "normal"
> operating conditions. Replacing the whole 1GB of memory would cost about
> £170, an expensive option. Where to start ? Software tests abound, but
> hardware diagnostics seem thin on the ground.

First place to start would be memtest. If that goes ok, then run Prime 95.
If that goes ok, I'd say you had a software problem with either the sound
card or tv card. Might try a Linux live CD like knoppix and see if it
crashes on you too.

--
KT133 MB, CPU @2400MHz (24x100): SIS755 MB CPU @2330MHz (10x233)
Need good help? Provide all system info with question.
My server http://wesnewell.no-ip.com/cpu.php
Verizon server http://mysite.verizon.net/res0exft/cpu.htm

 

[Guest]

Archived from groups: alt.comp.periphs.mainboard.abit (More info?)

I have an old ATI pci video/tv card that does this (system lock ups)
with winxp. Every driver/software version I have tried, over the past
5 years, for the card has had various forms of major defects. I
haven't bought anything else from ATI since.

One thing to try would be to remove the tv card and see if your
problems go away.

Forrest

Motherboard Help By HAL web site:
http://home.comcast.net/~mobo.help/


On Sun, 11 Sep 2005 19:41:12 +0100, Tom Goodman <tovmas@hotmail.com>
wrote:

< snip >
>.... If digital tv card is operative, then I notice CPU %
>fluctuate and creep up to 100% in windows task manager, then the lock
>up. sound sometimes "stutters" and the system locks up.
< snip >

 

[Guest]

Archived from groups: alt.comp.periphs.mainboard.abit (More info?)

Wes Newell wrote:
> On Sun, 11 Sep 2005 19:41:12 +0100, Tom Goodman wrote:
>
>
>>I have a further problem that I've not mentioned because I don't know
>>what the cause is and I'm reluctant to post in the wrong group. The
>>whole system "locks up" and neither keyboard or mouse respond. Reset is
>>necessary. If digital tv card is operative, then I notice CPU %
>>fluctuate and creep up to 100% in windows task manager, then the lock
>>up. sound sometimes "stutters" and the system locks up. (SB Live! card).
>>No overheating according to MBM, all drivers up-to-date and I suspect
>>hardware, but don't know how to start looking really. Sapphire Radeon
>>9200SE, DXdiag reports all ok. I've two 512MB memory modules and I
>>wonder of I should start by suspecting them ? Perhaps take one out at a
>>time and swap slots around. Memtest seems a recommended program, but I
>>see it can indicate faults which might not actually occur under "normal"
>>operating conditions. Replacing the whole 1GB of memory would cost about
>>£170, an expensive option. Where to start ? Software tests abound, but
>>hardware diagnostics seem thin on the ground.
>
>
> First place to start would be memtest. If that goes ok, then run Prime 95.
> If that goes ok, I'd say you had a software problem with either the sound
> card or tv card. Might try a Linux live CD like knoppix and see if it
> crashes on you too.
>
Thanks for that, Wes, I have knoppix and I'll see waht that does.
Before that I'll give memtest and prime 95 a blast.
Ran chkdsk /f and fixed an error relating to non-paged areas being paged.
Have "cleaned up" registry with Ashampoo, tweak now, and ace
utilities-hoping I've not blatted anything i shoulden't have.
Defragmented "C" completely as to space and going to run tests now.

 

[Guest]

Archived from groups: alt.comp.periphs.mainboard.abit (More info?)

Wes Newell wrote:

> First place to start would be memtest. If that goes ok, then run Prime 95.
> If that goes ok, I'd say you had a software problem with either the sound
> card or tv card. Might try a Linux live CD like knoppix and see if it
> crashes on you too.
>

Dear Wes,

Following your advice I've booted into Memtest and run 6 hours of
iterations of the test suite, followed by two complete iterations of the
fading bit test, and I'm glad to say that no errors were reported ! The
expense of replacing 1GB of memory seems less likely.

I don't yet have Prime95 but that will be the next thing to try.

After that I'll give Knopix a whiz, though I doubt it will recognise the
TV card or have a program for it.

I'll also be uninstalling software which I hardly ever use just in case
that might be causing problems.

I'm suspecting the Radeon 9200SE card, although I'm using the latest
catalyst drivers. I'm inclined to spend a bit on a better video card,
Nvidia based for a complete change and see if that will help. Problem is
the Radeon has no fan so is silent. I'm not sure that there is a fanless
Nvidia based card suitable for this motherboard voltage and I DON'T want
NOISE-I live in a quiet flat and I hate intrusive machinery. I've gone
to a quietPC power supply for that reason

Tom.

 

[Guest]

Archived from groups: alt.comp.periphs.mainboard.abit (More info?)

Wes,
Your advice produced results rapidly: Prime95 failed in less than one
minute using a variety of configurations !
I'm now inclined to suspect the CPU.
There has been no noticeable extra deterioration since overclocking the
Thorton from 2000MHZto 2266MHZ, and upping the bus speed from 100MHZ to
133MHZ.
The cheapest first option will be to replace the overclocked "Thorton
2400" with something that will run at 2266MHZ without overclocking. If I
can now get a suitable CPU of course.
Would you have a recommendation ?

Once again,
THANKS A LOT.

Tom

 

[Guest]

Archived from groups: alt.comp.periphs.mainboard.abit (More info?)

Wes,
Looking at a reply to an earlier poster, it seems that a mobile AthlonXP
would be better than an AthlonXP 3200+.
Because the internal multiplier is not locked. (Hard or not possible to
unlock the XP multiplier).

Would I be correct in thinking that the XP-M 2600+ would be the better
chio, with 133/266 FSB as standard and no multiplier locking ?

Tom.

 

[Guest]

Archived from groups: alt.comp.periphs.mainboard.abit (More info?)

Wes,
Looking at earlier posts I've ordered up an Athlon XP-M 2600+ from
overclockersUk, and I think that will fit the bill. Of course, it's
expensive compared to the USA price, £94.00 including postage +
insurance; USA price was about $75 excl, the lowest I could find.


I'm looking forward to testing the rig with this chip in !

Tom.

 

[Guest]

Archived from groups: alt.comp.periphs.mainboard.abit (More info?)

On Thu, 15 Sep 2005 19:31:59 +0100, Tom Goodman wrote:

> Wes,
> Looking at a reply to an earlier poster, it seems that a mobile AthlonXP
> would be better than an AthlonXP 3200+.
> Because the internal multiplier is not locked. (Hard or not possible to
> unlock the XP multiplier).
>
> Would I be correct in thinking that the XP-M 2600+ would be the better
> chio, with 133/266 FSB as standard and no multiplier locking ?
>
If you need performance of a socket a 2266Mhz, you'd be better off selling
your current MB/cpu combo and going to a socket 754 Sempron or A64 2800.

--
KT133 MB, CPU @2400MHz (24x100): SIS755 MB CPU @2330MHz (10x233)
Need good help? Provide all system info with question.
My server http://wesnewell.no-ip.com/cpu.php
Verizon server http://mysite.verizon.net/res0exft/cpu.htm

 

[Guest]

Archived from groups: alt.comp.periphs.mainboard.abit (More info?)

Wes Newell wrote:
> On Thu, 15 Sep 2005 19:31:59 +0100, Tom Goodman wrote:
>
>
>>Wes,
>>Looking at a reply to an earlier poster, it seems that a mobile AthlonXP
>>would be better than an AthlonXP 3200+.
>>Because the internal multiplier is not locked. (Hard or not possible to
>>unlock the XP multiplier).
>>
>>Would I be correct in thinking that the XP-M 2600+ would be the better
>>chio, with 133/266 FSB as standard and no multiplier locking ?
>>
>
> If you need performance of a socket a 2266Mhz, you'd be better off selling
> your current MB/cpu combo and going to a socket 754 Sempron or A64 2800.
>
Well, I would not feel happy selling a "dud" CPU, and apart from the
current problem I'm quite happy at the moment with the KT7A. So I think
if the Mobile XP works it's the better option. I'd also need to buy
another 1GB of ram since the PC133 would hardly do in a more modern board.

If the replacement CPU doesn't cure the bug(s) I'll consider the board
dud probably, then I'll look for another one to take the existing
CPU/ram and hardware. ANY SUGGESTION ?

Actually I have built an alternative system based on MSI KT6 Delta, with
Athlon XP 3500+ and 1GB of memory. In due course that will become my
main machine, hardware and software transferred to it, and the KT7A will
be retired. The only thing stopping me is that I have had several
ocurrences of XP Pro failing to boot because files have become corrupted
or missing ! I have had to do several recoveries and reinstalls and have
not got to the bottom of the problem yet. Twin Seagate Barracuda 250GB
7200 seem to be ok, memory tests ok. I suspect XP...but haven't finished
erxamining.

At present the KT6 box has a Suse Linux 9.3 Pro installed by ftp, and it
doesen't seem to suffer the same problem as it did under XP pro. It
boots every time and doesn't seem to suffer file corruption-or if it
does the system recovers itself automatically. I therefore strongly
suspect that for some reason the XP pro system/drivers can't handle this
mobo/ram/drives combination.

And further, under XP I can update the mobo drivers etc; with Linux I
can't because the checking/updating programs are Microsnot Windoze only
! So even with this limitation Linux has proved far superior. If I could
run the Nebula DigiTV card under Suse, I would think of dropping
windoze-although I have many programs I'm loath to part with.

Best wishes,
Tom.

 

[Guest]

Archived from groups: alt.comp.periphs.mainboard.abit (More info?)

On Fri, 16 Sep 2005 00:23:31 +0100, Tom Goodman wrote:

> If the replacement CPU doesn't cure the bug(s) I'll consider the board
> dud probably, then I'll look for another one to take the existing
> CPU/ram and hardware. ANY SUGGESTION ?
>
There aren't many boards left that will take non ddr ram. You migt see if
you can find a K7S5A or one of the boards using a KT266A chipset. Either
way though, you're limted to 2 sdram on these. If you have more than 2
sdram sticks, you're stuck with an older chipset like the KT133A you have
now.

--
KT133 MB, CPU @2400MHz (24x100): SIS755 MB CPU @2330MHz (10x233)
Need good help? Provide all system info with question.
My server http://wesnewell.no-ip.com/cpu.php
Verizon server http://mysite.verizon.net/res0exft/cpu.htm

 

[Guest]

Archived from groups: alt.comp.periphs.mainboard.abit (More info?)

Wes Newell wrote:
> On Fri, 16 Sep 2005 00:23:31 +0100, Tom Goodman wrote:
>
>
>>If the replacement CPU doesn't cure the bug(s) I'll consider the board
>>dud probably, then I'll look for another one to take the existing
>>CPU/ram and hardware. ANY SUGGESTION ?
>>
>
> There aren't many boards left that will take non ddr ram. You migt see if
> you can find a K7S5A or one of the boards using a KT266A chipset. Either
> way though, you're limted to 2 sdram on these. If you have more than 2
> sdram sticks, you're stuck with an older chipset like the KT133A you have
> now.
>
Thanks for those tips Wes.
I expected difficulty, but I'll go the extra mile to keep my 1GB (two
sticks) and the Mobile CPU now ordered. Otherwise I'm into a whole u/g
since My Radeon 9200SE is suitable only for the old AGP voltage I
believe, etc etc.

Tom.