On Oct 9, 2008, at 6:56 AM, Al wrote:

> By the technical standards of RAM design, there is no such thing as PC
> RAM and Mac RAM.

Correct.

RAM is RAM is RAM (to paraphrase Gertrude Stein, my Godmother).


> Things go awry when vendors do not comply with the
> technical design standards.

NO!

Things go awry when Apple's designers do not design their DRAM  
controller chips to support industry-standard JEDEC RAM sticks.



> That is how some RAM gets to be dirt
> cheap.

No.

Remember well: RAM is RAM is RAM.

RAM becomes dirt cheap when it becomes a commodity ... when it  
becomes so ubiquitous that it will work in any product (read: any PC).

RAM becomes dearly priced when it becomes machine specific ... when  
it becomes so specialized that it will work in only a few products  
(read: some, but not all Macs; etcetera).



> Regrettably, one may learn of an incompatibility on a specific
> Mac with a specific version of the OS only by trial, unless you buy
> from a vendor who certifies, or at least claims, that the RAM will
> meet the design standards assumed by Apple.  If you are looking at
> used RAM from PCs, or RAM that is otherwise non-specific, you get into
> the trial and error mode.
>

Why don't people understand that it is Apple which is the root cause  
of these incompatibilities, and that it is the end-user which is  
exacerbating these incompatibilities by seeking to buy cheap,  
commodity RAM for use in Apples which are misdesigned at the outset,  
and actually require proprietary RAM, without fully informing the end- 
user of the weaknesses in Apple's own designs.

RAM sticks are made in high density and low density; in one bank, two  
bank and four bank; in single-sided and double-sided. (Add 2K refresh  
and 4K refresh to the mix, for early PCI Macs)

Apple's DRAM controller chips only accept a subset, possibly a small  
subset of all these RAM sticks.

In general, only the last model in a series can accommodate the  
highest complexity RAM sticks, whereas the first model in a series  
can only accommodate the lowest complexity RAM sticks. Models in the  
middle may accommodate the medium complexity RAM sticks, or they may  
not. Only Apple knows for sure, and Apple isn't talking.


> This is a old issue.  I do not believe you can isolate the RAM sticks
> that will not work in a Mac by simply looking at its manufacturer's
> specifications.

Actually, you can, but you have to have access to literally tons of  
information on every possibly applicable RAM chip, and on the  
specific trade-offs employed in the printed circuit board designs.

Basically, the only folks who can do this right are the Apple VARs  
with their own in-house testing equipment.

A possible good example is OWC, but even an otherwise fine VAR like  
OWC has been known to fudge the specs and try to squeeze a silk purse  
out of a sow's ear.

Example 1: the so-called "converted" SRAM for early PCI Macs (7200,  
7500/7600, 7300, 8500/9500 and 8600/9600), which used higher density  
DRAM chips than Apple's DRAM controller could accommodate, but  
attempted to make this higher density DRAM appear to be the required  
lower density. These "converted" sticks would work on some PCI Macs,  
but not all, or in some sub-models, but not others.

Example 2: the so-called "S" SDRAM for mid-G4 Macs, which attempted  
to meet BOTH the PC100 specs for 100 MHz bus Macs, and the PC133  
specs for 133 MHz bus Macs. As Apple, in some cases, had as many as  
seven revisions of the LSI which controlled the RAM in these models,  
some examples would work with this RAM, but others would not. The  
obvious solution was to buy pure PC100 RAM for 100 MHz bus machines  
and pure PC133 RAM for 133 MHz bus machines, and to reject the RAM  
which tried to be all things to all machines. Some of the problems  
were related to changes in the LSI chips for dual processors, but  
others were related to the differences in 100 MHz buses and 133 MHz  
buses.


> Non-compliance with the design standard goes deeper
> than that.  I hope some experts on this list can weigh in with the
> technical details, such as I have seen as recently as a couple years
> ago.  Or you may find previous discussions in the archives.
>
> Even when third-party RAM is known to work on a Mac with a particular
> version of the OS, users have occasionally discovered that RAM to be a
> problem when attempting to upgrade the OS.  Through the years, Apple
> has demanded less and less tolerance for RAM production deviations.


When Apple began to experience more and more apparent "problems" ...  
all of its own doing, but for which it refused to accept  
responsibility ... Apple began to demand more and tighter RAM specs,  
and its later Mac OSes refused to work with RAM which did not meet  
the later, and more demanding RAM specs.

But, RAM is RAM is RAM, and Apple never fully appreciated that fact.

It had only to review and fully understand the industry-standard  
JEDEC specs, and then to design its DRAM controller LSIs to those specs.

That's all!

But, Apple didn't, and that was the origin of the problem.

Now that Apple is Intel-driven (in the sense that Intel is writing  
more of the Mac's specifications, down to the LSI chip level ... the  
MCH chip in particular) it is Intel and its MCH specifications which  
is defining RAM specifications, and Intel is more compliant with the  
JEDEC specs than Apple ever was.



--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed Low End Mac's G3-5 List, a 
group for those using G3, G4, and G5 desktop Macs - with a particular focus on 
Power Macs.
The list FAQ is at http://lowendmac.com/lists/g-list.shtml and our netiquette 
guide is at http://www.lowendmac.com/lists/netiquette.shtml
To post to this group, send email to [email protected]
To unsubscribe from this group, send email to [EMAIL PROTECTED]
For more options, visit this group at 
http://groups.google.com/group/g3-5-list?hl=en
Low End Mac RSS feed at feed://lowendmac.com/feed.xml
-~----------~----~----~----~------~----~------~--~---

Reply via email to