The dormant holographic storage volcano has rumbled, reminding us that it's not extinct and may yet erupt. As a reminder, this optical storage format came hot on the heels of previous storage successes: compact disks, magneto-optical drives, DVDs, and then Blu-ray. Blu-ray never reached the same relative scale of popularity as …
All I want
All I want is a way to store my data in such a way that 10, 20, or more years from now I know I will be able to retrieve it. Of course, it would have to be priced appropriately for the consumer market.
It is not unusual for writable CDs, DVDs, and BDs to go bad after even a few months.
Stone slabs, chisels.
Proven storage lifetime of the order of tens of thousands of years. Low capital investment, low ongoing costs. Reading and writing equipment has remained pretty much the same since its first release.
Write speeds and capacities are rather low of course, but you didn't specify those requirements in detail so I guess this'll do just fine.
lintel on top
Ah, back to the age of the silicon chunk..
[ acks to TPratchett ]
Re: All I want
"10, 20, or more years from now I know I will be able to retrieve it"
How about a RAID array of something cheap-per-megabyte? (I'm not sure if optical discs actually *are* cheap per megabyte compared with bottom-end external hard discs.)
"How about a RAID array of something cheap-per-megabyte?"
Repeat to yourself over and over, RAID is *not* a backup solution and is *not* an archive.
that in your eagerness to chant the mantra you've missed the rather obvious point that a RAID array makes a perfectly decent backup target.
Speed also not important
This tech is taking an age to come of ages but I think we need some alternative to hard drives for backup. And if the killer app is cheap long term storage it has to be reliable and cheaper (taking into account the burner and disks) than hard drives but it could be much slower. 30 hours for a 500Gb cheap backup would be fine for most people for home use.
Have to disagree.
Any backup which requires more than 24 hours to create is doomed to failure.
This will fail too - I don't use optical disks for archive purposes, I use HDD's. They are cheap enough at sufficient sizes (+1TB) that I can simply buy one every year or two and move the data to the new one and use the old one in a live machine.
The rolling nature of the HDD's means the drive is always useable on modern hardware, I get ever increasing storage capacity on my network and I can keep my data updated and organised.
Only threat I have now is in the slow evolution of formats over time - I still have some ancient documents I wrote in the mid-90's with a 3rd party office suit the name of which time has claimed from my memory. PF-somthing-something... I know it wasn't IBM's.
Have you tried with Libreoffice? (Not affiliated with them in any way, just impressed by the number of formats supported)
the drive is always useable on modern equipment...
I take it you mean writable...
What do you plug this year's equivalent of the 5 1/2 in MFM winchester into in 20 yrs time?
Will noone rid us of these fragile media?
Spinning optical disks fail unless stored and used in hermetically sealed vaults, and even then they may have nktnhave burned correcty. Is anyone developing a way to store data cheply but securely in a format that doesn't fail if it's dropped on a slightly rough carpet?
The data storage technology of the future. Always has been, probably always will be.
Will the disc melt before write is finished?
If I recall, these lasers put out a chunk of heat - such that good ripping software (EAC) has a 10-min cool-down pause per hour.
Writing probably uses more power than reading. 3 or 30 hours (depending on speed) -- they'll need serious cooling to ensure you don't end up with a pile of holographic plastic slag dripping out of the drive!
(flame because it's... well... hot!)
AC had it right
The threat is not that your precious bits will get lost. At least not the _main_ threat. Rolling from one generation of HDD to another works fine for that.
The threat is that when you try to open your PageMaker4 or Word3.0 files, you will be darn lucky to get anything recognizable back. Similarly for any other application (nearly all of them) that uses a proprietary binary storage format.
Unless there is an EMP
then all magnetic media is toast.
Optical media heading for the scrapheap
Spinning media is good in one way. It requires no power to store. However if you have a large archive it takes a bit more effort to keep it organised and then you have to go digging for it. HDD's are good and RAID arrays are better. However it takes power to keep it up and running and drive failures are a fact of life. Also we are now having to swtich from IDE to SATA as the interface gets phased out on modern motherboards.
I am looking forward to SSD's becoming larger and cheaper. They take less power, cause less heat and should be much more durable than spinning media. Also RAID should be needed less. Since SSD's being used for storage requires few writes, they should theoretically last a lifetime (assuming there is hardware to read them).
Eventually though we will have a small matchbox sized unit for local storage and everything else stored in the 'cloud'.
Micro-Hologram is not the only architecture
Thanks Chris for bringing this subject some much needed attention. There has been a bit of a "quiet" period in this space, but rest assured, there is exciting development being done in the labs and research facilities to advance this technology. I posted a response to your blog here,
I describe what Hitachi is cooking with Holographic storage and how it compares with GE's architecture. Cheers.