Questions about DDR RAM

In alt.comp.hardware.pc-homebuilt (Nate Edel) wrote:

In comp.sys.ibm.pc.hardware.chips krw wrote:
In article ,
says...
In comp.sys.ibm.pc.hardware.chips krw wrote:
Memory is 64bits wide these days. Run your numbers again.

Quibble: *DIMMs* are 64-bits wide these days.

We were talking about memory.

Yes; all DIMMs are memory, but not all memory comes in DIMMs.

Memory comes in lots of widths, and no more explicitly applies to DIMMs than
to the individual chips on the DIMMs or soldered to a board.

Memory, even just talking about main memory on the PC architecture, runs
from 64 bits to 256 (possibly 512, if you count AMD's lightweight NUMA as
parallel. Otherwise, 256 is for the quad-channel Xeon boards.)

We were talking about memory, not processor implementation. Do you
have an example of an x86 processor with a wider than 64bit data path
to the DIMM (ignoring dual channel for the moment, which is simply
two DIMMs)?

x86? No, although you're changing the goalposts, and I did say it was a
quibble. The AMD dual channel implementation *IS* a full 128-bit path to
memory, just split between two DIMMs.

I have no idea what IBM uses for memory in their very high-end stuff, but
that's not x86; even high-end stuff from Sun, etc, tends to just use
multiple channels breaking down to DIMMs.

ISTR some fairly ridiculous bus widths for high-end graphics cards, as well.

We were... Do you really have parity/ECC on graphics cards?

Not on consumer stuff.

Not running code, but pretty much only data, there really isn't much
*need* for ECC on graphics cards, as long as the error isn't obvious and
permanent. (There are tests for that.) Who cares all that much if a
pixel changes color slightly?

It's been a *long* time since I've paid attention to high-end workstation
graphics boards. It would be pretty implausible on consumer stuff or on
today's "derived from consumer" FireGL or Quadro boards.

Anyone have an old IBM PGA board from the mid-80s? I'd be quite surprised if
that DIDN'T have parity.

--
_____
/ ' / â„¢
,-/-, __ __. ____ /_
(_/ / (_(_/|_/ / _/ _

In article ,
says...
In comp.sys.ibm.pc.hardware.chips krw wrote:
In article ,
In comp.sys.ibm.pc.hardware.chips krw wrote:
In article ,
In comp.sys.ibm.pc.hardware.chips krw wrote:
Memory is 64bits wide these days. Run your numbers again.

Quibble: *DIMMs* are 64-bits wide these days.

We were talking about memory.

Yes; all DIMMs are memory, but not all memory comes in DIMMs.

Memory comes in lots of widths, and no more explicitly applies to DIMMs than
to the individual chips on the DIMMs or soldered to a board.

Oh, good ****ing grief Gert! I know the Democrats in the Senate
don't believe it, but CONTEXT MATTERS. How long have DIMMs been the
only memory used?

The only memory used for what? Main-memory on mainstream PC motherboards,
it's probably close to ten years since 72-pin SIMMs went out.

In PCs, dipstick! Has another DRAM form factor been used in the last
decade for anything (other than directly mounted)?

Other uses? Never. There are plenty of other sorts and uses of memory.

Examples...

They may or may not use ECC. That wasn't the point I was making.

I'm not sure what point you're making anymore, but you're looking
rather foolish dancing.

Memory, even just talking about main memory on the PC architecture, runs
from 64 bits to 256 (possibly 512, if you count AMD's lightweight NUMA as
parallel. Otherwise, 256 is for the quad-channel Xeon boards.)

We were talking about memory, not processor implementation. Do you
have an example of an x86 processor with a wider than 64bit data path
to the DIMM (ignoring dual channel for the moment, which is simply
two DIMMs)?

x86? No, although you're changing the goalposts, and I did say it was a
quibble. The AMD dual channel implementation *IS* a full 128-bit path to
memory, just split between two DIMMs.

No, I'm not moving goal posts (look at the name of the NG). You're
being an pedantic ass.

A) This thread is being posted to several newsgroups:
alt.comp.hardware.hardware.homebuilt
alt.comp.hardware
alt.comp.hardware.pc-homebuilt
comp.sys.ibm.pc.hardware.chips
... and in practice, the last (where I'm reading) has plenty of
broader-ranging discussions.

....and those groups are involved with?

B) Second, see above quoting my original post: I started by noting that it
was, indeed a quibble. You're the one objecting to my clarification,
which was a minor point. If that's "being an pedantic ass", deal with it.

Dance, dance, dance.

I have no idea what IBM uses for memory in their very high-end stuff, but
that's not x86; even high-end stuff from Sun, etc, tends to just use
multiple channels breaking down to DIMMs.

I do know. Registered DIMMs, in fact. Anything else would be
stupid. Your point is?

So the z-series and similar uses DIMMs? Fair enough.

Yes. Anything else would be sorta pointless, no?

ISTR some fairly ridiculous bus widths for high-end graphics cards, as well.

We were... Do you really have parity/ECC on graphics cards?

Not on consumer stuff.

Not on *ANY* stuff. Can you think of *one* reason to spend a nickel?
Add the complication?

Raw framebuffers have no need for parity/ECC. Graphics memory these days is
hardly just a raw framebuffer.

Dance some more. References to this ECC graphics memory, please!

It's been a *long* time since I've paid attention to high-end workstation
graphics boards. It would be pretty implausible on consumer stuff or on
today's "derived from consumer" FireGL or Quadro boards.

Anyone have an old IBM PGA board from the mid-80s? I'd be quite surprised if
that DIDN'T have parity.

I'd be *very* surprised if it did (in fact I know they didn't). I
have *NEVER* seen parity on graphics memory. WTF would anyone waste
money on that? What is the gain, one off-color pixel? Come on, stop
running away from your silliness.

So it only ran out of the ROM routines?

Huh? You've danced right off the edge!

--
Keith