🚀 go-pugleaf

In article <896785475.27116.0.nnrp-07.9e98c909@news.demon.co.uk>
           henry@halon.demon.co.uk "HAL" writes:

> Is there any software available that can distinguish between a 66 or 100MHZ
> DIMM ? Is there any physical difference between 66 and 100MHZ SDRAM ?

Memory is specified not in the CPU's clock speed but by access time 
in nanoseconds. 

If you want advice on what access speeds are appropriate for 
different models and speeds of CPU and with numbers of wait states 
set, you will need to ask somwehere other than in demon.ip.support.pc 
which deals only with ip software running under DOS.

Copied & fu set, including demon.tech.pc.

Also, please never start a thread which cross-posts between groups 
unless you clearly understand the charters of each, as even the 
helpful replies will continue to pollute the groups that should never 
have been included in the posting.  In newsgroups, guessing the scope 
of a charter from the name of the group is a BAD THING to do.

Regards,
-- 
   Tony - G3SKR / W2TG                   email: tgold@panix.com

Firstly demon.ip.support.pc in an inappropriate group for this question, I
have copied this followup to demon.tech.pc and set followups accordingly.

In article <896785475.27116.0.nnrp-07.9e98c909@news.demon.co.uk>, HAL
<henry@halon.demon.co.uk> writes
>Is there any software available that can distinguish between a 66 or 100MHZ
>DIMM ? 

The only way to be sure, is to look at the chips (usually 8 or 16 chips of a
54pin TSOP II package). Take down the full details, then go hunting the
manufacturers Websites looking for the chips datasheet. To give you an idea, I
have included a list below of chips in Intels PC100 spec. testing document.
This should help you find the datasheet you require. Also hunt around and find
the Intel Document "PC100 SDRAM Component Testing Summary" which details what
is required in order for a chip to be PC100 compliant.

Note that you have to know a bit about memory in order to read datasheets,
otherwise they are 100pages of gibberish. Try researching in Toms Hardware
guide (www.tomshardware.com) and Anands hardware guide (www.anandtech.com).

Other than that, you have to take your suppliers word for it.

>Is there any physical difference between 66 and 100MHZ SDRAM ?

If there is a little 8pin chip (24C02 or simular), then that is the SPD or
Serial Presence Detect chip. All PC100 SDram modules should have one of those,
but many PC66 modules do too.

Note that 32Mb Dimms are usually made up of 16 of the 16Mbit chips (16 chips
of 8 bits wide is 64bits wide, and 16 chips of 2Mb is 32Mb) though some are
made with 4 64Mbit (4M x 16).

64Mb Dimms are usually made with 8 of the 64Mbit (8M x 8) chips.


Take care, I hope that this helps,


Mark.........
-------------------------------------- --------- ----------------------------
Supplier Device Part Number/Date       Code      Comment
-------------------------------------- --------- ----------------------------
Fujitsu 16M - 2Mx8 MB81F16822B102FNE   S9745-M06 under evaluation
Hyundai 16M - 2Mx8 HY57V168010C TC-10S 9801TA    pass all parameters tested
LG 16M - 2Mx8 GM72V16821DT-7K          9809      pass all parameters tested
Mitsubishi 16M - 2Mx8 M5M4V16S30DTP-8  811       pass all parameters tested
NEC 16M - 2Mx8 D4516821AG5-A10-7JF     9743BY018 under evaluation
Samsung 16M - 2Mx8 KM48S2020CT-GL ES   807       pass all parameters tested
Siemens 16M - 2Mx8 HYB39S16800AT-8     C9736     under evaluation
TI 16M - 2Mx8 TMX626812BDGE5M          82A1TO P  pass all parameters tested
Toshiba 16M - 2Mx8 G52805 9644MBD C59S1608AFT-10 under evaluation
-------------------------------------- --------- ----------------------------
Fujitsu 64M - 8Mx8 81F64842B-103FN     9750 K01R1 pass all parameters tested
Hitachi 64M - 8Mx8 HM5264805TT-B60     9748      pass all parameters tested
Hyundai 64M - 8Mx8 HY57V658020ALTC-10P9749TA K0021E pass all parameters tested
LG 64M - 8Mx8 GM72V66841CT-7J          9748      pass all parameters tested
Mitsubishi 64M - 8Mx8 M5M4V64S30ATP-8  806       pass all parameters tested
NEC 64M - 8Mx8 D4564841G5-A10-9JF      9812      pass all parameters tested
Samsung 64M - 8Mx8 KM48S8030BT-GH      801       pass all parameters tested
Siemens 64M - 8Mx8 39S64800AT          9812      pass all parameters tested
TI 64M - 8Mx8 TMX664814A81A7ET                   pass all parameters tested
Toshiba 64M - 8Mx8 A56877 9750KBD TC59S6408BFT-80 pass all parameters tested
-------------------------------------- --------- ----------------------------
Fujitsu 64M - 4Mx16 81F641642B-103FN   9805 MOO  pass all parameters tested
Hitachi 64M - 4Mx16 HM5264165TT-B60    9751      pass all parameters tested
Hyundai 64M - 4Mx16 HY57V651620ATC-10P 9803      pass all parameters tested
LG 64M - 4Mx16 GM72V661641CT-7J        9806      pass all parameters tested
Mitsubishi 64M - 4Mx16 M5M4V64S40ATP-8 810       pass all parameters tested
NEC 64M - 4Mx16 D4564163G5-A10-9JF     9803E9001 under evaluation
Samsung 64M - 4Mx16 KM416S4030BT-G10   FS 807    pass all parameters tested
Toshiba 64M - 4Mx16 TC59S6416BFT-80              pass all parameters tested
-------------------------------------- --------- ----------------------------

(Sorry these lines are 78chars long, the table wouldnt fit otherwise)
-- 
Mark Booth (markbSig@oxim.demon.co.uk)
--------------------------------------------------
Oxford Intelligent Machines Ltd, Oxford, UK.
Voice +44 (0) 1865 204881, Fax +44 (0) 1865 204882

On Tue, 02 Jun 98 12:05:55 GMT, tgold@microvest.demon.co.uk ("Anthony
R. Gold") wrote:

>In article <896785475.27116.0.nnrp-07.9e98c909@news.demon.co.uk>
>           henry@halon.demon.co.uk "HAL" writes:
>
>> Is there any software available that can distinguish between a 66 or 100MHZ
>> DIMM ? Is there any physical difference between 66 and 100MHZ SDRAM ?
>
>Memory is specified not in the CPU's clock speed but by access time 
>in nanoseconds. 

  Actualy - for SDRAM this isn't strictly true.  Most of the
datasheets show the speed in MHz.  Sometimes a 'ns equivalent' value
is shown as well.

  The difference between these units is speed (obviously enough), but
to determine this for a random DIMM is hard unless you have a special
test unit.   Why don't the markings on the chips suffice?

   Regards,
          Harry.