In need of utillity to identify Chipset

  • Thread starter Thread starter JBJ
  • Start date Start date
J

JBJ

Hi, I have now tried with Astra32, SIW, Everest and Intels chipset utility
tool but all of them give different results and Intels tool give alot of
different chipsets.

So I'm looking for a tool that will give the correct chipset of my laptop.

I thinks its 945GM but I have to be sure.

Also if someone know a tool that will tell which socket is the laptop:)

Thanks:)
 
JBJ said:
Hi, I have now tried with Astra32, SIW, Everest and Intels chipset
utility tool but all of them give different results and Intels tool give
alot of different chipsets.

So I'm looking for a tool that will give the correct chipset of my laptop.

I thinks its 945GM but I have to be sure.

Also if someone know a tool that will tell which socket is the laptop:)

Thanks:)
Click to expand...

http://www.cpuid.com/cpuz.php

The socket type is a physical attribute. In a laptop, there is only
room for certain kinds of sockets. You can use the following link, to
learn what options are available for the various Intel processors.

http://processorfinder.intel.com

Paul
 
Paul said:
http://www.cpuid.com/cpuz.php

The socket type is a physical attribute. In a laptop, there is only
room for certain kinds of sockets. You can use the following link, to
learn what options are available for the various Intel processors.

http://processorfinder.intel.com

Paul
Click to expand...


Hi Paul, thanks for the answer, unfortunately CpuZ cant make a precise
answer either since it claims that it is i943/940GML. Its not both, so i
still need a programme that will tell me excactly what is in the laptop:)

A supporttech from Toshiba(actually a seperate company doing support for
Toshiba) claimed it was 943GML but I cant get that to be true with the
drivers installed.

Toshiba takes forever to come out with new driver where Intel allready have
them available, as long as I know what chipset it is.

Unfortunately Intels chipset util says this:

Detected Chipset:
Mobile Intel(R) 945GM/GMS/940GML Express Chipsets or Intel(R) 945GT Express
Chipset

Not very clear about what it precisely is, but it does not say 943GML.

I dont have alot of confidense in the supporttech since there is alot he
should know that he doesn't, dont get me wrong he is a nice guy but it seems
that he is not really technical and that he just reads what it says on the
screen.
 
JBJ said:
Hi Paul, thanks for the answer, unfortunately CpuZ cant make a precise
answer either since it claims that it is i943/940GML. Its not both, so i
still need a programme that will tell me excactly what is in the laptop:)

A supporttech from Toshiba(actually a seperate company doing support for
Toshiba) claimed it was 943GML but I cant get that to be true with the
drivers installed.

Toshiba takes forever to come out with new driver where Intel allready
have them available, as long as I know what chipset it is.

Unfortunately Intels chipset util says this:

Detected Chipset:
Mobile Intel(R) 945GM/GMS/940GML Express Chipsets or Intel(R) 945GT
Express Chipset

Not very clear about what it precisely is, but it does not say 943GML.

I dont have alot of confidense in the supporttech since there is alot he
should know that he doesn't, dont get me wrong he is a nice guy but it
seems that he is not really technical and that he just reads what it
says on the screen.
Click to expand...

The datasheet for the chipset is here. If you look for "Device Identification"
in the document, not all the chipsets can be separated by having a
different enumeration. That could account for CPUZ not providing an
answer. It appears one differentiator, is a different clock operating
speed for the graphics core in the Northbridge.

http://download.intel.com/design/mobile/datashts/30921905.pdf

I think I'd start with the CPUZ "About" tab, and dump the registers.
That will give you a cpuz.txt file.

For example, in the datasheet, section 5.1.33, it talks about the
CAPID register contents. Perhaps that can be used to figure out
which chip it is.

Paul
 
Paul said:
The datasheet for the chipset is here. If you look for "Device
Identification"
in the document, not all the chipsets can be separated by having a
different enumeration. That could account for CPUZ not providing an
answer. It appears one differentiator, is a different clock operating
speed for the graphics core in the Northbridge.

http://download.intel.com/design/mobile/datashts/30921905.pdf

I think I'd start with the CPUZ "About" tab, and dump the registers.
That will give you a cpuz.txt file.

For example, in the datasheet, section 5.1.33, it talks about the
CAPID register contents. Perhaps that can be used to figure out
which chip it is.

Paul
Click to expand...

Hi Paul, I have not had much time for this lately so excuse my late
response. I tried getting the information again, from Toshiba but hit a
brick wall, the boss of the support tech. Appearently its not often they get
this kind of questions, so he is reluctant to give it to me, but finally he
answered my question in a not nice way(snotty). Unfortunately he keeps
repeating what the tech did and read it from papers instead of looking at
the machine itself. Papers can be wrong.

His response were "If the customer knows so much about chipset then he must
know that ....."

Anyway he claims that it is a 943GML chipset with a 943GML graphic chip, and
the reason for it saying it is a 945GM/GU is that all the 940 - 945 chipset
use the same driver.
Then he refers to this page:
http://www.intel.com/support/chipsets/sb/CS-022034.htm
And to this: http://www.intel.com/design/chipsets/mobile/945_fam.htm

so his response is that eventhough Intels graphic driver shows a 945GM/GU
graphic chip then its really a 943GML. And the reason for showing 945
instead of 943 is that its a generic driver.

I tried calling intel but they only support their own motherboards and
refered back to Toshiba.

So basically I hit a brick wall.
 
JBJ said:
Hi Paul, I have not had much time for this lately so excuse my late
response. I tried getting the information again, from Toshiba but hit a
brick wall, the boss of the support tech. Appearently its not often they
get this kind of questions, so he is reluctant to give it to me, but
finally he answered my question in a not nice way(snotty). Unfortunately
he keeps repeating what the tech did and read it from papers instead of
looking at the machine itself. Papers can be wrong.

His response were "If the customer knows so much about chipset then he
must know that ....."

Anyway he claims that it is a 943GML chipset with a 943GML graphic chip,
and the reason for it saying it is a 945GM/GU is that all the 940 - 945
chipset use the same driver.
Then he refers to this page:
http://www.intel.com/support/chipsets/sb/CS-022034.htm
And to this: http://www.intel.com/design/chipsets/mobile/945_fam.htm

so his response is that eventhough Intels graphic driver shows a
945GM/GU graphic chip then its really a 943GML. And the reason for
showing 945 instead of 943 is that its a generic driver.

I tried calling intel but they only support their own motherboards and
refered back to Toshiba.

So basically I hit a brick wall.
Click to expand...

If you browse over the Intel datasheet, I think you'll come to the same
conclusion. In terms of Device_IDs coming from the device, there is little
to differentiate the chipsets. The datasheet covers a family of devices
which is based on the same silicon. They speed up the clock inside the
chip, when the chip is used in a desktop. They slow it down, for the
laptop cases. The internal clock is synthesized - something like a 66MHz
input clock, is multiplied up to whatever clock is desired. (At least,
that is what is used on my current computer, as a reference clock.)

The driver installed, would likely be based on matching the Device_ID. So
if you want to call it a "generic" driver, then in that sense, the
concept is correct. But once installed, it should still be possible for
the driver to provide info on what exactly the chip is.

Using the CAPID, and looking at the capabilities, that tells you which
of the sub-cases it might be. But a normal driver installer wouldn't be
looking for something like that.

I would blame this on Intel, because if Intel wanted, they could have
come up with a scheme to make the Device_ID controllable by external
straps. So whatever mechanism they use to set the capabilities, could
be reflected in a different Device_ID.

Which is why I'm suggesting looking at things like the CAPID, as the
clock choice being used should match a laptop case. Is is still possible
that the CAPID is not exact enough, to declare an exact match.

And have a look through the datasheet. I may have missed something
besides the Capabilities register.

Another differentiator, might be the interfaces supported in the design.
Like if it has a VGA connector, a DVI connector, HDMI, S-video, component,
composite, and so on. For a manufacturer like Intel, it is easy to
enable or disable things like that at the factory, then change the
part number on the top of the chip. Then, the question would be,
from a software point of view, can the output options be identified.
And that is why you should go through the datasheet.

Paul
 
Paul said:
If you browse over the Intel datasheet, I think you'll come to the same
conclusion. In terms of Device_IDs coming from the device, there is little
to differentiate the chipsets. The datasheet covers a family of devices
which is based on the same silicon. They speed up the clock inside the
chip, when the chip is used in a desktop. They slow it down, for the
laptop cases. The internal clock is synthesized - something like a 66MHz
input clock, is multiplied up to whatever clock is desired. (At least,
that is what is used on my current computer, as a reference clock.)

The driver installed, would likely be based on matching the Device_ID. So
if you want to call it a "generic" driver, then in that sense, the
concept is correct. But once installed, it should still be possible for
the driver to provide info on what exactly the chip is.

Using the CAPID, and looking at the capabilities, that tells you which
of the sub-cases it might be. But a normal driver installer wouldn't be
looking for something like that.

I would blame this on Intel, because if Intel wanted, they could have
come up with a scheme to make the Device_ID controllable by external
straps. So whatever mechanism they use to set the capabilities, could
be reflected in a different Device_ID.

Which is why I'm suggesting looking at things like the CAPID, as the
clock choice being used should match a laptop case. Is is still possible
that the CAPID is not exact enough, to declare an exact match.

And have a look through the datasheet. I may have missed something
besides the Capabilities register.

Another differentiator, might be the interfaces supported in the design.
Like if it has a VGA connector, a DVI connector, HDMI, S-video, component,
composite, and so on. For a manufacturer like Intel, it is easy to
enable or disable things like that at the factory, then change the
part number on the top of the chip. Then, the question would be,
from a software point of view, can the output options be identified.
And that is why you should go through the datasheet.

Paul
Click to expand...


Hi Paul, thanx for the answer, but you might as well be speaking russian:))
Could you give me a 1,2,3 guide to how to do that you just descrived. The
dataheet I read but its mumbo jumbo to me these days(maybe it will come
clear to me in a month while doing something else, hehehehe) I had an
accident a few years back and sometimes my brain doesn't work like i want it
to, and then again sometimes my family calls me a genius:-).
 
JBJ said:
Hi Paul, thanx for the answer, but you might as well be speaking
russian:)) Could you give me a 1,2,3 guide to how to do that you just
descrived. The dataheet I read but its mumbo jumbo to me these
days(maybe it will come clear to me in a month while doing something
else, hehehehe) I had an accident a few years back and sometimes my
brain doesn't work like i want it to, and then again sometimes my family
calls me a genius:-).
Click to expand...

This is a problem in pattern matching, as what you're doing, is
comparing the info in the datasheet, to the tables dumped in CPUZ.txt.

Where I'd start, is with the register dump you can get from CPUZ. In the
"About" tab, there is an option to dump the registers. In one of the
hex tables of numbers, would be a copy of the current contents
of the memory space of the Northbridge. CAPID is going to be in
there.

For example, crack open 30921905.pdf and turn to page 78 which has
Table 3. Notice there are "default values" listed. "Capability Identifier"
is at offset 0xE0 to 0xE8, and would be 72 bits total in all. What I'd
do, is pattern match the values I see on page 78, to what gets dumped
by CPUZ. If I find a match in one of the tables, then I know I'm
looking at the same stuff documented in the Intel datasheet.

The values seen in CPUZ are byte swapped. The value "8086 27A0" or
"8086 27A6" at the beginning of Device 0 Configuration, would show
as "86 80 A0 27" or "86 80 A6 27", and that is how I'd tell I was
in the right table.

Section 5.1.33 on page 107 defines CAPID. CAPID is defined in terms
of bits, so you copy the 9 bytes from the table and rewrite it
as 72 bits. Then read the definitions of the bit fields. For
example, bits 62, 61, and 60 are supposed to identify the
sub type.

001: Mobile Intel 945GM/GME Express Chipset
010: Mobile Intel 945GMS/GU Express Chipset
110: Mobile Intel 943/940GML Express Chipset
011: Mobile Intel 945PM Express Chipset
101: Intel® 945GT Express Chipset

Bits 31, 30, 29 define FSB maximum speed support. Apparently there
is a difference between 943GML and 940GML, that the 940 only
supports FSB533.

011: (G)MCH capable of up to FSB 667
100: (G)MCH capable of up to FSB 533

If you want help, post the contents of "CPUZ.txt", which
is generated by doing a "registers dump" using the "About" tab
in CPUZ.

You can snip the section near the end, starting with "DMI BIOS" down to the end.
That will save a few lines.

The section "Thread dumps" at the beginning can be removed.
The section "Memory SPD" up to "Monitoring" can be removed.

By doing that, the remainder of the information in the file
I have here, is slightly below 32KB and may fit in one
USENET post.

Paul
 
Back
Top