Overclocking the PB 600 MB.

Hello Bob,,,,

The more I remember what you mentioned:

Wait, you said you're running a 100MHz CPU, right ?

That means you are already at 66MHz FSB. 1.5X66MHz=
100MHz


And, comparing this with the 600 MB specs from the following:

http://www.uktsupport.co.uk/pb/mb/600.htm

It apears that I do have the 66MHz CPU,,,,, If I remember correctly,
the process I ran reports 100MHz CPU processor Speed and no L1 and/or
L2. Plus, the front panel mentions 100MHz.

But, the specifications reports L1 and L2 Cach as well as the 75/133
CPU in the following list:

Bus Architecture : PCI 2.0/ISA based system bus.
Cache : 16KB L1 (Internal) write-back Cache (integrated in the CPU).
256KB L2 (External) write-through with write buffers cache
soldered on the motherboard.
Chipset :
PCI064B Super I/O controller.
VL82C593 controller.
CPU : Pentium 75 MHz Pentium 133 MHz.Uses Type 5 CPU Zero Insertion
Force (ZIF) CPU Socket.

So, this specification apears to be for another CPU brand and speed (
a 75, 133, MMX, ????) with L1 and L2. And, these specs do not describe
what I actually have relative to the cpu, L1 andL2 cache(????)..

And, it looks like all the remiauing entries in the specs can be
matched up with what I am seeing on the board. Even the picture of the
board indicates the same layout of parts on the board.


Thanks for your time, advice, and help!!!

Jerry

On Mon, 01 Oct 2007 19:15:14 GMT, Jerry wrote:

Hello agin BOB,,,,,

Sorry,,,I meant to mention this in the previous entry-----

Even though the board is out when I put in a PCI NIC (BTW: it works),
I did not look directly at these jumpers as they are configured on the
board. The system is running good and responsive!

OK, about the J32 jumper on 2X requires a 133,,,and would slow down a
100,,,,,,, THANKS for that warning,,,,, and no matter how they are
set I'm gonna leave them as they are set.

Jerry


On Mon, 01 Oct 2007 19:02:32 GMT, Jerry wrote:

Hello Bob,,,,


OK,,,, and based on the Jumper settings in the specs as default with
the astric (*) mark these are the following:

J30 Host Bus Frequency * In 60/66 MHz
Out 50 MHz

J31 Host Bus Frequency In 66 MHz
* Out 50/60 MHz

J32 Host Core Freq. Multiplier * In 2 X
Out 1.5X


As far as overclocking or speeding up this CPU, it apears it came that
way from the factory.......

However, I do not know what the J31 jumper controls?

And, I had made the last entries at the bottom in this Newsgroup
THREAD before I found this one from you.


THANKS for your time, help, and advice!!!!

Jerry

On Fri, 28 Sep 2007 18:04:39 -0400, "Robert E. Watts"
wrote:

Hello Jerry !

Wait, you said you're running a 100MHz CPU, right ?

That means you are already at 66MHz FSB. 1.5X66MHz= 100MHz

And no, the difference is not at all eyepopping, just better. 8=)

I haven't studied your jumpers, the description is easy to find on the
internet. In any event, the only ones you can play with are the frequency
jumpers ( whichever that is. )

Don't bother changing the multiplier. That CPU will run slower, but it won't
run at 2X. You will need a 133, 150, 166, or 200 CPU for that. See my
previous message.

bob



"Jerry" wrote in message
...
Hello Bob,,,,

OK,,,,I'll give the J31 a try and set it to 66MHz.

Based on your comments it must be an "eye-popping" or at least some
visually noticable difference,,,,,especially in I/O ~~~~(PCI, Video,
and Memory I/O) .

This looks like a case of---- "so what,,,,it ain't gonna hurt to set
it to 66MHz",,,,, and there will probably be some visual notice of
increased performance,,,If so or If not,,,then so be it!

Would this include the the I/O for the IDE/ATAPI interface for
disk,tape, cd, and other related I/O devices (floppy????) ?

Based on your description I hope I am beginning to see more about
"What's What" in this PB 600 MB specification for these jumpers.

It appears the J31 "Host Bus Freqeuncy" is the "Front Side Bus
Frequency" which affects I/Ocomponens on the MB, etc. on the MB.

What does the jumper J30 "Host Bus Freqeuncy" affect?

And, the J32 "Host core freq. Multiplier" looks like it affects the
speed of the CPU-----(overclocking????).

Looking up the term "Host Bus Frequency" on the the Internet was too
many to read but the ones that were read did not help to identiify
what specifically was affected. I would think that the availablity of
these type jumpers on a motherboard vary from one MB to another as
well as what specifcally is provided by these type jumpers which may
also vary from MB to MB (????).

Please excuse me if these are stupid questions and statements,,,
correct me if the statement is incorrect ,,,,the last time I delt wth
disk I/O and tape I/O, etc. was before the term DMA was on the
horizon,,,,,

THANKS,,,,,,

Jerry

Fri, 28 Sep 2007 06:55:40 -0400, "Robert E. Watts"
wrote:

HI Jerry !

( comments below )

"Jerry" wrote in message
om...
Hello Bob,,,,,,


OK,,,,,I just dug through my records that I had printed out from this
Group relating to ~~~600 Upgrades~~~. That was a long time ago and I
had forgotten I had these. I am surprised I still had them (vintagee
Group dat 1997-2003). And, I apologize for any inconvience.

Based on the dicussion here and those other listings along with the
pin orientaion listed on the specs for the 600MB, I now see the
default delivery for that board is set for max cpu as you mentioned
below. Next time I get a look at the MB I will make sure these are the
same. They should be because it was purchased new "off the shelf".

The pin orientation for J31 is set default as Host Bus Frequency at
50/60 Mhz and I doubt that setting it to 66Mhz would show any
"eye-ball popping" noticable increase in performance.



Actually, when you set the FSB ( front side bus ) speed to 66MHz, as
opposed
to 50 or 60MHz, you are not just increasing the speed of the CPU. Even
more
importantly, you are "speeding" up the motherboard. The I/O access is
faster, PCI slot ( video too! ) is faster ( max speed of 33Mhz), and
memory
read/write speed is faster. This is a very noticeable increase in speed.
And
66Mhz is not really "overclocking". Almost all components are designed to
handle this speed.

Aftermarket clone boards had settings of 75 and 83MHz ( and some had even
more settings, sometimes in 1MHz increments ). This would REALLY speed
things up, but sometimes created problems with some components. Certain
CPU's like the Cyrix were designed to run at 75MHz, but other components (
cheap ones usually ) had problems with this.

Then of course the Super Socket 7 boards came along, with speeds of 100Mhz
and beyond.

Intel practically never allows "overclocking" on their boards, and since
I'm
not a big overclocker, I never cared. I prefer Intel motherboards ( now )
over almost everything else. Simply the most stable boards there are ( in
my
opinion ). And PB used a lot of Intel boards.

I read that you mention your board has 256K of L2 cache. Good for you.
This
is about the best you can do when it comes to a PB board, and it really
helps. Does your case say "Platinum" on it? My *research* indicates that
after Socket 4 boards, only the Platinum designated PB computers have L2
cache. I have yet to see one with 512k.

bob

Hi Jerry !

Real quick ( 'cause the Bengals are on Monday Night Football tonight. Not
that I am a big Bengals fan, but I'm a HUGE Monday Night Football fan.
Packer Backer here ! )

L1 cache is built into the CPU ( in this case, we are referring to old Intel
Pentium 1 CPU's )

Socket 5 and 7 Intel CPU's all have 16K of L1 cache. The MMX versions
have 32K.
( this is another reason that AMD CPU's were faster back then. They
had more, sometimes as much as 256K on the K6-3 ! )

L2 cache is on the motherboard. Packard Bell computers usually don't have
it. I have not seen a Platinum version that did not have L2. Doesn't mean
they all have it, but my empirical evidence so far points to this fact.

I have not seen L2 cache on other PB units. This does not take into account
a motherboard change. A PB board that has L2 cache is usually 256k. I have
seen mention of PB boards with 512k, but no one that I know of has seen one.
Maybe Ben has.

You can easily look on the board, and see if it has L2 or not. Also, it is
reported in the BIOS screen on boot up.

bob

Hi Jerry !

Download this program:

http://www.cpuid.com/


.... and run it. You will find out lots of interesting information about your
computer instantly.

:-)

bob

Hello Bob,,,,

OOPPSSS ,,,, sorry the dicussion about CACHE is different than the
"Overclocking" !!!

OK and thanks for the link to CPU-Z----- turns out I had the 1.33
version and had forgotten about it.

I got the 1.41 version that is now available. I have already used it
on several computers today and it apears that if the cpuz .exe does
not report CACHE the latency.exe will if it is found.

I did find that 1.41 cpuz.exe will not run on W95B. Gets error:

Link to missing export Kernel32.DLL:IsDebuggerPresent

And an error for latency.exe:

Link to missing export Kernel32.DLL:SetProcessAffinityMask

If available for W95B, I may need the Kernel32.DLL that has these
entry point/logic for W95B. Otherwise, 1.41 runs ok on W98SE.
And 1.33 runs OK on W95B as well as W98SE. So this makes it not
necessary even if one is available for W95B. And, 1.33 reports cache
if found.

THANKS!!!!

Jerry



On Mon, 1 Oct 2007 18:25:10 -0400, "Robert E. Watts"
wrote:

Hi Jerry !

Real quick ( 'cause the Bengals are on Monday Night Football tonight. Not
that I am a big Bengals fan, but I'm a HUGE Monday Night Football fan.
Packer Backer here ! )

L1 cache is built into the CPU ( in this case, we are referring to old Intel
Pentium 1 CPU's )

Socket 5 and 7 Intel CPU's all have 16K of L1 cache. The MMX versions
have 32K.
( this is another reason that AMD CPU's were faster back then. They
had more, sometimes as much as 256K on the K6-3 ! )

L2 cache is on the motherboard. Packard Bell computers usually don't have
it. I have not seen a Platinum version that did not have L2. Doesn't mean
they all have it, but my empirical evidence so far points to this fact.

I have not seen L2 cache on other PB units. This does not take into account
a motherboard change. A PB board that has L2 cache is usually 256k. I have
seen mention of PB boards with 512k, but no one that I know of has seen one.
Maybe Ben has.

You can easily look on the board, and see if it has L2 or not. Also, it is
reported in the BIOS screen on boot up.

bob

Hello Bob,,,,

OK,,,just loaded CPU-Z 141 and rebooted:

The BIOS reports:

256K CACHE SRAM passed

CPU-Z 141 reports:

CPU
NAME: Intel Pentium
Code Name: P54
Package: Socket 5(296)
Core Speed bounces back and forth showing 99.5 and 99.6 MHz

CACHE

L1 D-Cache
Size: 8 KBytes
Descriptor: 2-way set associative, 32-byte line size

L1 I-Cache
Size: 8 KBytes
Descriptor: 2-way set associative, 32-byte line size


NO INFORMATION reported for L2 Cache

CPU-Z 141 Latency.exe reports the same.

The other process that is run to ID MB stuff is called Belarc and it
reports no Cache at all.

So, it looks like, I think, what you had mentioned previously ---

16KB CACHE (????) (default delivered with this CPU and 600 MB)

Is there some explanation for the BIOS reported 256 and what CPU-Z
reports? Or, am I misinterpreting or not understanding what the BIOS
reported?

Looks like I do not understand what the BIOS is actually reporting
with the following display during BOOT:

256K CACHE SRAM passed


THANKS for your time, help, and advice!!!!

Jerry




On Tue, 02 Oct 2007 17:50:09 GMT, Jerry wrote:

Hello Bob,,,,

OOPPSSS ,,,, sorry the dicussion about CACHE is different than the
"Overclocking" !!!

OK and thanks for the link to CPU-Z----- turns out I had the 1.33
version and had forgotten about it.

I got the 1.41 version that is now available. I have already used it
on several computers today and it apears that if the cpuz .exe does
not report CACHE the latency.exe will if it is found.

I did find that 1.41 cpuz.exe will not run on W95B. Gets error:

Link to missing export Kernel32.DLL:IsDebuggerPresent

And an error for latency.exe:

Link to missing export Kernel32.DLL:SetProcessAffinityMask

If available for W95B, I may need the Kernel32.DLL that has these
entry point/logic for W95B. Otherwise, 1.41 runs ok on W98SE.
And 1.33 runs OK on W95B as well as W98SE. So this makes it not
necessary even if one is available for W95B. And, 1.33 reports cache
if found.

THANKS!!!!

Jerry



On Mon, 1 Oct 2007 18:25:10 -0400, "Robert E. Watts"
wrote:

Hi Jerry !

Real quick ( 'cause the Bengals are on Monday Night Football tonight. Not
that I am a big Bengals fan, but I'm a HUGE Monday Night Football fan.
Packer Backer here ! )

L1 cache is built into the CPU ( in this case, we are referring to old Intel
Pentium 1 CPU's )

Socket 5 and 7 Intel CPU's all have 16K of L1 cache. The MMX versions
have 32K.
( this is another reason that AMD CPU's were faster back then. They
had more, sometimes as much as 256K on the K6-3 ! )

L2 cache is on the motherboard. Packard Bell computers usually don't have
it. I have not seen a Platinum version that did not have L2. Doesn't mean
they all have it, but my empirical evidence so far points to this fact.

I have not seen L2 cache on other PB units. This does not take into account
a motherboard change. A PB board that has L2 cache is usually 256k. I have
seen mention of PB boards with 512k, but no one that I know of has seen one.
Maybe Ben has.

You can easily look on the board, and see if it has L2 or not. Also, it is
reported in the BIOS screen on boot up.

bob

Hello Jerry !

If the BIOS reports 256K CACHE SRAM, that is what you have.

Don't know why CPU-ID isn't finding it.

Interesting, and good for you (that you have it. :-)

bob




"jglong3" wrote in message
...
Hello Bob,,,,


The BIOS reports:

256K CACHE SRAM passed

CPU-Z 141 reports:

CPU
NAME: Intel Pentium
Code Name: P54
Package: Socket 5(296)
Core Speed bounces back and forth showing 99.5 and 99.6 MHz

CACHE

L1 D-Cache
Size: 8 KBytes
Descriptor: 2-way set associative, 32-byte line size

L1 I-Cache
Size: 8 KBytes
Descriptor: 2-way set associative, 32-byte line size


NO INFORMATION reported for L2 Cache

CPU-Z 141 Latency.exe reports the same.

The other process that is run to ID MB stuff is called Belarc and it
reports no Cache at all.

So, it looks like, I think, what you had mentioned previously ---

16KB CACHE (????) (default delivered with this CPU and 600 MB)

Is there some explanation for the BIOS reported 256 and what CPU-Z
reports? Or, am I misinterpreting or not understanding what the BIOS
reported?

Looks like I do not understand what the BIOS is actually reporting
with the following display during BOOT:

256K CACHE SRAM passed


THANKS for your time, help, and advice!!!!

Jerry




On Tue, 02 Oct 2007 17:50:09 GMT, Jerry wrote:

Hello Bob,,,,

OOPPSSS ,,,, sorry the dicussion about CACHE is different than the
"Overclocking" !!!

OK and thanks for the link to CPU-Z----- turns out I had the 1.33
version and had forgotten about it.

I got the 1.41 version that is now available. I have already used it
on several computers today and it apears that if the cpuz .exe does
not report CACHE the latency.exe will if it is found.

I did find that 1.41 cpuz.exe will not run on W95B. Gets error:

Link to missing export Kernel32.DLL:IsDebuggerPresent

And an error for latency.exe:

Link to missing export Kernel32.DLL:SetProcessAffinityMask

If available for W95B, I may need the Kernel32.DLL that has these
entry point/logic for W95B. Otherwise, 1.41 runs ok on W98SE.
And 1.33 runs OK on W95B as well as W98SE. So this makes it not
necessary even if one is available for W95B. And, 1.33 reports cache
if found.

THANKS!!!!

Jerry



On Mon, 1 Oct 2007 18:25:10 -0400, "Robert E. Watts"
wrote:

Hi Jerry !

Real quick ( 'cause the Bengals are on Monday Night Football tonight. Not
that I am a big Bengals fan, but I'm a HUGE Monday Night Football fan.
Packer Backer here ! )

L1 cache is built into the CPU ( in this case, we are referring to old
Intel
Pentium 1 CPU's )

Socket 5 and 7 Intel CPU's all have 16K of L1 cache. The MMX
versions
have 32K.
( this is another reason that AMD CPU's were faster back then. They
had more, sometimes as much as 256K on the K6-3 ! )

L2 cache is on the motherboard. Packard Bell computers usually don't have
it. I have not seen a Platinum version that did not have L2. Doesn't mean
they all have it, but my empirical evidence so far points to this fact.

I have not seen L2 cache on other PB units. This does not take into
account
a motherboard change. A PB board that has L2 cache is usually 256k. I
have
seen mention of PB boards with 512k, but no one that I know of has seen
one.
Maybe Ben has.

You can easily look on the board, and see if it has L2 or not. Also, it
is
reported in the BIOS screen on boot up.

bob

Some of the newer CPU ID programs do not have the necessary algorithms to detect
an older level 2 cache separate from the CPU itself... Ben Myers

On Wed, 3 Oct 2007 18:11:39 -0400, "Robert E. Watts" wrote:

Hello Jerry !

If the BIOS reports 256K CACHE SRAM, that is what you have.

Don't know why CPU-ID isn't finding it.

Interesting, and good for you (that you have it. :-)

bob

Boy, what memories of my PB days. Just stopped by and found familiar
territory. The Powerleap/PB 600 combination was one that was recognized
by the 600's Phoenix BIOS. Had a K6-2 clocked to 400 that ran like a
top until I couldn't find a small enough HDD replacement! Time marches on.

Ben Myers wrote:
The OP would have to make sure that the Powerleap kit includes the necessary
BIOS modifications to allow the BIOS to recognize the non-Intel CPU... Ben Myers

On Sat, 29 Sep 2007 05:52:14 -0700, "mdp" wrote:

Jerry wrote:
Hello,,,,,,

The following instructions have been found to speed up
(overcloking????) the PB 600 MB 100MHZ Pentium (Ithink????):

1. J30 closed
2. J31 closed
3. J32 Closed
4. J28 the two inside pins connected


Would like to understand the following:

1. What speed should be expected?
2. Will this require a cooling fan?

THANKS!!!!!

Jerry
You might consider a Powerleap adapter and a K6-2 400MHz CPU (if a K6-3
works the difference will be dramatic). Some other upgrade stories he

http://www.geocities.com/siliconvall...ory/pb600.html

Other general info he

http://www.geocities.com/SiliconVall...774/pb600.html

http://www.geocities.com/SiliconVall.../8774/cpu.html

Hello All,,,

OK,,,, I must admit this 1995 vintage cpu is rather old to say the
least!!!! After using Belarc, CPU-Z, and RightMarks's 3.72 it does
look like this is the case based on what Ben mentions below.

I am now using the 600 to do this entry to the Group. This is the
first time in a day or so that I have had the chance to work with this
MB plus with the specifications list "in hand" at the same time. So,
with this convience, I am looking directly on the MB and see the chip
sets labeled U68-U71 and U78-U81 on the specifications list and they
are populated on the MB.

This appears to be "proof positive" the CACHE is there(???).

And now that I am using this MB, "System Properties" reports 88MB RAM.
The MB is populated with the 8MB on-board memory with 4 available
slots for other memory,,,, I have sticks in all 4 to include the one
8MB factory installed stick(16MB as delivered). I installed the
others to get up to 88MB and that was an "eye-popping" increase in
performance.


THANK you all for your comments, time, and advice!!!!

Jerry

On Wed, 03 Oct 2007 21:16:04 -0400, Ben Myers
wrote:

Some of the newer CPU ID programs do not have the necessary algorithms to detect
an older level 2 cache separate from the CPU itself... Ben Myers

On Wed, 3 Oct 2007 18:11:39 -0400, "Robert E. Watts" wrote:

Hello Jerry !

If the BIOS reports 256K CACHE SRAM, that is what you have.

Don't know why CPU-ID isn't finding it.

Interesting, and good for you (that you have it. :-)

bob

Hello All,,,,,

And, based on what has been discussed here and in the THREAD "CACHE",
I am fairly sure of what is actually configured on this MB and that it
is running about as good as it can run with the stuff currently on the
MB. Plus, I continue to be satisfied with the performance, especially
now that it has 88MB RAM. That was a real "eye opener" increase in
performance.

THANKS to all for your comments, time, and advice!!!

Jerry


On Wed, 03 Oct 2007 21:01:53 -0700, Kryos wrote:

Boy, what memories of my PB days. Just stopped by and found familiar
territory. The Powerleap/PB 600 combination was one that was recognized
by the 600's Phoenix BIOS. Had a K6-2 clocked to 400 that ran like a
top until I couldn't find a small enough HDD replacement! Time marches on.

Ben Myers wrote:
The OP would have to make sure that the Powerleap kit includes the necessary
BIOS modifications to allow the BIOS to recognize the non-Intel CPU... Ben Myers

On Sat, 29 Sep 2007 05:52:14 -0700, "mdp" wrote:

Jerry wrote:
Hello,,,,,,

The following instructions have been found to speed up
(overcloking????) the PB 600 MB 100MHZ Pentium (Ithink????):

1. J30 closed
2. J31 closed
3. J32 Closed
4. J28 the two inside pins connected


Would like to understand the following:

1. What speed should be expected?
2. Will this require a cooling fan?

THANKS!!!!!

Jerry
You might consider a Powerleap adapter and a K6-2 400MHz CPU (if a K6-3
works the difference will be dramatic). Some other upgrade stories he

http://www.geocities.com/siliconvall...ory/pb600.html

Other general info he

http://www.geocities.com/SiliconVall...774/pb600.html

http://www.geocities.com/SiliconVall.../8774/cpu.html