Mass production of revolutionary computer memory moves closer with ULTRARAM TM on silicon wafers for the first time...

[server]

On Sat, 8 Jan 2022 19:44:14 +0200, Dimiter_Popoff <dp@tgi-sci.com>
wrote:

On 1/8/2022 18:38, jlarkin@highlandsniptechnology.com wrote:
On Fri, 7 Jan 2022 23:08:44 +0200, Dimiter_Popoff <dp@tgi-sci.com
wrote:

On 1/7/2022 22:25, John Larkin wrote:
On Fri, 7 Jan 2022 12:12:15 -0800 (PST), whit3rd <whit3rd@gmail.com
wrote:

On Friday, January 7, 2022 at 10:20:41 AM UTC-8, John Larkin wrote:

Universities churn out revolutionary breakthroughs weekly, or maybe
daily. To the nearest 0.1%, none work out.

Not a sensible description of what happens. Universities aren\'t claiming
\'revolutionary\' in general, just reporting progress. The dramatic descriptions come
from news reporters, and popularizers, and promoters, who are all
non-academic folk whose affiliations are with the entertainment press.

Most universities have a press department that hypes these miracles.


Well probably so but if this delivers not 1000 years (as extrapolated),
just 100 years of data retention it will be groundbreaking.
If (which is a biggish if of course) this works this will spell the
end of magnetic storage as we know it (and it will be a shame, today\'s
disks are really great pieces of engineering).

Hard drives are great. But keep backups. RAID is good too.

I backup on DVD-s, apart from more than one HDD (and more than one
partition on these). In the 80-s, when all I had were 8\" floppies
on the least reliable drives I have seen (sort of a Bulgarian clone
of some I think)
I learnt the lesson that a single backup is not enough...
I am saying this with my fingers crossed of course, you never know.

We dump all the company server files onto a fresh USB hard drive,
every two months. Those are treated as write-once devices, never to be
written to and probably never to be even connected to a computer
again. A 2 Tbyte drive costs around $60 now.

There are basically no floppy drives around any more. I wonder how
long USB ports will appear on PCs, until we go all wireless. Probably
a long time.

There are semiconductor storage chips that are (somehow) rated for
10,000 years retention. A usb memory stick in a freezer will probably
do that.


Somehow I can\'t make myself trust these tiny charges. Not that I don\'t
rely on them, I use flash and I did use eproms etc. all of my life
while designing systems but here I am.

--

I yam what I yam - Popeye

 

[Dimiter_Popoff]

On 1/8/2022 20:23, jlarkin@highlandsniptechnology.com wrote:

On Sat, 8 Jan 2022 19:44:14 +0200, Dimiter_Popoff <dp@tgi-sci.com
wrote:

On 1/8/2022 18:38, jlarkin@highlandsniptechnology.com wrote:
On Fri, 7 Jan 2022 23:08:44 +0200, Dimiter_Popoff <dp@tgi-sci.com
wrote:

On 1/7/2022 22:25, John Larkin wrote:
On Fri, 7 Jan 2022 12:12:15 -0800 (PST), whit3rd <whit3rd@gmail.com
wrote:

On Friday, January 7, 2022 at 10:20:41 AM UTC-8, John Larkin wrote:

Universities churn out revolutionary breakthroughs weekly, or maybe
daily. To the nearest 0.1%, none work out.

Not a sensible description of what happens. Universities aren\'t claiming
\'revolutionary\' in general, just reporting progress. The dramatic descriptions come
from news reporters, and popularizers, and promoters, who are all
non-academic folk whose affiliations are with the entertainment press.

Most universities have a press department that hypes these miracles.


Well probably so but if this delivers not 1000 years (as extrapolated),
just 100 years of data retention it will be groundbreaking.
If (which is a biggish if of course) this works this will spell the
end of magnetic storage as we know it (and it will be a shame, today\'s
disks are really great pieces of engineering).

Hard drives are great. But keep backups. RAID is good too.

I backup on DVD-s, apart from more than one HDD (and more than one
partition on these). In the 80-s, when all I had were 8\" floppies
on the least reliable drives I have seen (sort of a Bulgarian clone
of some I think)
I learnt the lesson that a single backup is not enough...
I am saying this with my fingers crossed of course, you never know.

We dump all the company server files onto a fresh USB hard drive,
every two months. Those are treated as write-once devices, never to be
written to and probably never to be even connected to a computer
again. A 2 Tbyte drive costs around $60 now.

Our \"serious\" data (i.e. designs, sources etc.) is on dps machines only
and a backup still fits on a DVD (would have no chance to do so if we
were using outside design and programming tools).
The USB HDDs are nice, we have several full of films and tv-shows
we thought we might re-watch (and practically never did/do.... but who
knows what is around the corner).
The photo archive has grown larger than a dvd, it goes on a few bluray
disks and on some of the usb HDD-s, plus some of it going to my
dps hdd-s.
I don\'t back the dps images to usb drives mainly due to a policy
\"no access to our designs for any alien OS\", not that I have been
that fixated on this but I managed to maintain it for decades so
why now drop it.

There are basically no floppy drives around any more. I wonder how
long USB ports will appear on PCs, until we go all wireless. Probably
a long time.

The last time I accessed some of our floppies was some 10-15 years
ago as I moved them to image files under dps (I also have an emulation
of the old 6809 based machine which used these floppies).
Interestingly some 10-20 year old floppies were readable but any attempt
to write to them was damaging the sector write was attempted to.

 

[whit3rd]

On Friday, January 7, 2022 at 11:36:25 PM UTC-8, Jan Panteltje wrote:

On a sunny day (Fri, 07 Jan 2022 10:20:30 -0800) it happened John Larkin
jlarkin@highland_atwork_technology.com> wrote in
t01htg1ik93cpn7ei...@4ax.com>:
On Fri, 07 Jan 2022 05:24:16 GMT, Jan Panteltje
pNaonSt...@yahoo.com> wrote:

Mass production of revolutionary computer memory moves closer with ULTRARAM TM on silicon wafers for the first time

https://onlinelibrary.wiley.com/doi/10.1002/aelm.202101103

Looks interesting, 1000 years data retention.

The pictures in the paper of the silly-con look very doable to me.
Sure we will have to wait for it to be in the local shop.

To entrust our data to it, we have to wait for stocking distributors; to pony up
investment, opportunity will arrive sooner.

The silicon parts are do-able, it\'s the OTHER exotic materials in the mix that are bothersome;
nano-sheets of indium arsenide and gallium antimonide and aluminum antimonide, and no internal strains
or cracks allowed, for a few gigabytes of data, will have to compete with DVD-R variants
for cost (unlikely this decade) and write speed (needs work) and packing/storing/retrieving
in some variant filing system (that, seems like it could happen).

Compared to flash, the size of bits is huge, and the process layers other than the silicon foundation
are relatively unknown territory.

 

[Jan Panteltje]

On a sunny day (Sat, 8 Jan 2022 12:05:23 -0800 (PST)) it happened whit3rd
<whit3rd@gmail.com> wrote in
<5fdfcfcf-9e1d-4acd-85a1-06c6c23f1c4dn@googlegroups.com>:

On Friday, January 7, 2022 at 11:36:25 PM UTC-8, Jan Panteltje wrote:
On a sunny day (Fri, 07 Jan 2022 10:20:30 -0800) it happened John Larkin
jlarkin@highland_atwork_technology.com> wrote in
t01htg1ik93cpn7ei...@4ax.com>:
On Fri, 07 Jan 2022 05:24:16 GMT, Jan Panteltje
pNaonSt...@yahoo.com> wrote:

Mass production of revolutionary computer memory moves closer with ULTRARAM TM on silicon wafers for the first time

https://onlinelibrary.wiley.com/doi/10.1002/aelm.202101103

Looks interesting, 1000 years data retention.

The pictures in the paper of the silly-con look very doable to me.
Sure we will have to wait for it to be in the local shop.

To entrust our data to it, we have to wait for stocking distributors; to pony up
investment, opportunity will arrive sooner.

The silicon parts are do-able, it\'s the OTHER exotic materials in the mix that are bothersome;
nano-sheets of indium arsenide and gallium antimonide and aluminum antimonide, and no internal strains
or cracks allowed, for a few gigabytes of data, will have to compete with DVD-R variants
for cost (unlikely this decade) and write speed (needs work) and packing/storing/retrieving
in some variant filing system (that, seems like it could happen).

Good point

Compared to flash, the size of bits is huge, and the process layers other than the silicon foundation
are relatively unknown territory.

Yes, for now I have backups on DVD / Bluray, USB harddisks (2) and USB FLASH.
Also some nice stuff on M-DISC (using an LG DVD burner), supposed to last a long time:
https://en.wikipedia.org/wiki/M-DISC
My advice with exactly one thousand DVDs in an alu box now: keep those stored in the absolute dark.
It is the same as with photography, exposure time * light intensity will kill your data.
In a day if on the bookshelf in a transparent plastic box in the sun.
Oldest CD-R discs now 20 years or more and still readable.
http://panteltje.com/pub/CD_box_binnenkant_IXIMG_0549.JPG