Question about terminology

[Scott]

I believe hard drives and tape are collectively known as 'magnetic
media' and CDs, DVDs and Blu-Ray are collectively know as 'optical
media'. Is there a generic term for memory sticks, SD cards and solid
state drives?

 

[Ken Blake]

I believe hard drives and tape are collectively known as 'magnetic
media'
Yes.

and CDs, DVDs and Blu-Ray are collectively know as 'optical
media'.
Yes.

Is there a generic term for memory sticks, SD cards and solid
state drives?

No, none that is commonly used.

 

[Andy Burns]

No, none that is commonly used.

Flash media? Some confusion with .flv files

Non-Volatile solid state media? Quite a mouthful.

 

[Ken Blake]

Flash media?

SD cards are called flash drives, or flash media, but not memory
sticks or SSDs.

 

[John Williamson]

I believe hard drives and tape are collectively known as 'magnetic
media' and CDs, DVDs and Blu-Ray are collectively know as 'optical
media'. Is there a generic term for memory sticks, SD cards and solid
state drives?

Flash media for the removable types. I'm not aware of a special term for
SSDs.

 

[BillW50]

Flash media for the removable types. I'm not aware of a special term for
SSDs.

They use the same chips actually. The only difference is the interface
and the controller. Back in the very early days they were called EEPROM
(electronic erasable programmable read only memory). The major
difference between EEPROM and flash memory is EEPROM is programmable
byte by byte vs. flash which is programmable by blocks (much faster).

 

[Yousuf Khan]

I believe hard drives and tape are collectively known as 'magnetic
media' and CDs, DVDs and Blu-Ray are collectively know as 'optical
media'. Is there a generic term for memory sticks, SD cards and solid
state drives?

They are usually composed of flash memory chips internally, so they are
often called flash media. But in the not-too-distant future, they maybe
replaced by newer types of non-volatile memory, such as MRAM or other
things, so an even more generic name for them could be just simply
non-volatile memory/storage.

Yousuf Khan

 

[John Williamson]

They are usually composed of flash memory chips internally, so they are
often called flash media. But in the not-too-distant future, they maybe
replaced by newer types of non-volatile memory, such as MRAM or other
things, so an even more generic name for them could be just simply
non-volatile memory/storage.

And as they're random access, then they could be called NVRAM. Now,
where have I heard that before?

 

[Stan Brown]

I believe hard drives and tape are collectively known as 'magnetic
media' and CDs, DVDs and Blu-Ray are collectively know as 'optical
media'. Is there a generic term for memory sticks, SD cards and solid
state drives?

They're EEPROMs, I believe, but I don't think anyone uses that term
any more.

 

[Yousuf Khan]

And as they're random access, then they could be called NVRAM. Now,
where have I heard that before?

They wouldn't be called NVRAM, because RAM usually is associated with
temporary storage, and it would cause too much confusion amongst
computer users, used to talking about RAM as non-permanent storage. NV
storage would be more likely to be used instead.

Yousuf Khan

 

[Dave \Crash\ Dummy]

I believe hard drives and tape are collectively known as 'magnetic
media' and CDs, DVDs and Blu-Ray are collectively know as 'optical
media'. Is there a generic term for memory sticks, SD cards and solid
state drives?

I don't know of any formal or official term. I refer to them all as
"solid state" media.

 

[Andy Burns]

They wouldn't be called NVRAM, because RAM usually is associated with
temporary storage, and it would cause too much confusion amongst
computer users, used to talking about RAM as non-permanent storage. NV
storage would be more likely to be used instead.

Watching one of the talks from LinuxConf.au by Matthew Wilcox of Intel,
that's the way some are expecting storage to go, rather than making
flash (or other non volatile technologies) look like a disc, just map it
directly into address space and let it stay put over a power loss, or
present it as a direct file system, rather than blocks on a pretend disc.

You could call it core storage.

http://mirror.linux.org.au/linux.conf.au/2013/ogv/The_future_of_nonvolatile_memory.ogv

 

[Char Jackson]

Watching one of the talks from LinuxConf.au by Matthew Wilcox of Intel,
that's the way some are expecting storage to go, rather than making
flash (or other non volatile technologies) look like a disc, just map it
directly into address space and let it stay put over a power loss, or
present it as a direct file system, rather than blocks on a pretend disc.

You could call it core storage.

http://mirror.linux.org.au/linux.conf.au/2013/ogv/The_future_of_nonvolatile_memory.ogv

Sounds like how the Commodore 64 (and others of the era) did it.

 

[Ken Blake]

You could call it core storage.

That wouldn't really be an appropriate name.

Back when I first started programming, in 1962 on an IBM 1401, the
1401 and all other computers (or almost all, if I'm not right about
"all") had core storage. Core storage was functionally the same as
when we now call RAM, but it was very different in structure. It was
made of tiny magnetic rings (called "cores," which is why the memory
was called "core storage") with wires threaded through them that
carried electrical currents. Depending on how the pulses went through
those wires, the cores were magnetized in one direction (1) or the
other (0) and could be read by other wires. So each core, which was
either 0 or 1, was a bit of storage.

The 1401 that I first worked on had 4K bytes (4,000, not 4096) of core
storage. Each byte consisted of seven bits, so 4,000 bytes used 28,000
of those cores. Tiny as the cores were, 28,000 was a lot of them, and
the 4,000 bytes was about the size of a modern refrigerator. Today,
RAM is all electronic and no longer uses those cores, so it isn't core
storage.

And today's electronic RAM is *much* smaller than core storage. At the
moment I have in my hand a 32GB thumb drive, which is a little smaller
than my thumb. That 32GB is 8,589,934 times as much memory as the 4K
on the 1401 (and other thumb drives that are the same physical size
can contain even more memory), and many thousands of time smaller.
It's also many thousands of times less expensive.

And to add another point, today's electronic memory is much faster
than core storage ever was. At least in part, that's because it's much
smaller than core storage. It takes electricity about one nanosecond
to traverse an 11 inch wire, so to achieve faster speeds, memory (and
other electronics) needs to be smaller.

When a program crashed, you printed out the contents of core storage
to aid in diagnosing what the problem was. That printout was called a
"core dump." The term "core dump" is still sometimes used for such
printouts of RAM, even though it's no longer cores that are dumped,
but as far as I know, the term "core storage" is no longer used.

 

[Andy Burns]

That wouldn't really be an appropriate name.
Back when I first started programming, in 1962 on an IBM 1401, the
1401 and all other computers (or almost all, if I'm not right about
"all") had core storage. Core storage was functionally the same as
when we now call RAM, but it was very different in structure. It was
made of tiny magnetic rings

[snip]

yes, of course we know that, the type of storage they're pondering in
the video isn't necessarily NAND flash where reads are a bit slow and
writes are a lot slow compared to RAM, and DMA transfers have to be done
in blocks of a few K, they're talking about possibly 50% the speed of
current RAM, possibly approaching the same speeds, and operating at the
level of a cache-line worth of data just like reading writing normal memory.

Now, I'm not old enough to have worked with core memory, the first time
I was shown some was in school in the mid 70's when it was already
"antique", but the way new non-volatile storage *could* end up being
used, doesn't sound so different from the way core was used to me?

 

[Mike Barnes]

Now, I'm not old enough to have worked with core memory, the first time
I was shown some was in school in the mid 70's when it was already
"antique", but the way new non-volatile storage *could* end up being
used, doesn't sound so different from the way core was used to me?

Core memory was usually working storage, analogous to RAM today. Yes,
the contents persisted without power, but that feature was not much used
IME, because core was too expensive for long-term storage. That's not to
say it was never used: I came across software being delivered on core
boards, where there was no other convenient medium.

 

[Andy Burns]

Core memory was usually working storage, analogous to RAM today. Yes,
the contents persisted without power, but that feature was not much used

Fair enough.

 

[Yousuf Khan]

I don't know of any formal or official term. I refer to them all as
"solid state" media.

Yes, Solid State Media or Solid State Storage might be another term
usable. The only problem with it is that it's too close to Solid State
Drives (SSD's). SSD's would be a subset of this, true, but other things
like USB thumb drives would also be put under this umbrella. So it might
confuse people when they are called Solid State Media or Storage, and
someone might confuse it with an SSD.

In the end, this might all be simply called "storage", as eventually
older forms of storage are no longer used, and it's no longer necessary
to distinguish them.

Yousuf Khan

 

[Mellowed]

That wouldn't really be an appropriate name.

Back when I first started programming, in 1962 on an IBM 1401, the
1401 and all other computers (or almost all, if I'm not right about
"all") had core storage. Core storage was functionally the same as
when we now call RAM, but it was very different in structure. It was
made of tiny magnetic rings (called "cores," which is why the memory
was called "core storage") with wires threaded through them that
carried electrical currents. Depending on how the pulses went through
those wires, the cores were magnetized in one direction (1) or the
other (0) and could be read by other wires. So each core, which was
either 0 or 1, was a bit of storage.

The 1401 that I first worked on had 4K bytes (4,000, not 4096) of core
storage. Each byte consisted of seven bits, so 4,000 bytes used 28,000
of those cores. Tiny as the cores were, 28,000 was a lot of them, and
the 4,000 bytes was about the size of a modern refrigerator. Today,
RAM is all electronic and no longer uses those cores, so it isn't core
storage.

And today's electronic RAM is *much* smaller than core storage. At the
moment I have in my hand a 32GB thumb drive, which is a little smaller
than my thumb. That 32GB is 8,589,934 times as much memory as the 4K
on the 1401 (and other thumb drives that are the same physical size
can contain even more memory), and many thousands of time smaller.
It's also many thousands of times less expensive.

And to add another point, today's electronic memory is much faster
than core storage ever was. At least in part, that's because it's much
smaller than core storage. It takes electricity about one nanosecond
to traverse an 11 inch wire, so to achieve faster speeds, memory (and
other electronics) needs to be smaller.

When a program crashed, you printed out the contents of core storage
to aid in diagnosing what the problem was. That printout was called a
"core dump." The term "core dump" is still sometimes used for such
printouts of RAM, even though it's no longer cores that are dumped,
but as far as I know, the term "core storage" is no longer used.

Wow. Did you bring back memories! I was only a Project Engineer at
that time (actually 1970 time frame) and we used HP mini computers
(weighed about 100lbs) with 4K core memories for both position
resolution using the Transit satellite system and to provide navigation
solution on TTY machines. Programing was Machine code to squeeze it in
the available memory. It paid the bills.

 

[Ken Blake]

Now, I'm not old enough to have worked with core memory, the first time
I was shown some was in school in the mid 70's when it was already
"antique", but the way new non-volatile storage *could* end up being
used, doesn't sound so different from the way core was used to me?

The way it was used wasn't different at all. As I said, "Core storage
was functionally the same as when we now call RAM, but it was very
different in structure."