Is Holographic Storage the future of archive & backup?

And now for something completely different.  GE researchers have announced that they have successfully demonstrated a micro-holographic material that can support 500 GB in a DVD-style disc.  That's 20 times greater than most Blu-Ray discs (there is a Blu-Ray 100 in the works), and 100 times greater than DVDs.  So does this have backup and archive potential?  Let's look into that.

The first question is how fast this thing will be.  The article said that it supports "data recording at the same speed as Blu-ray discs."  The fastest a Blu-Ray disc can currently write is 12x, which translates into 54 MB/s.  That's slow in comparison to modern tape drives, but still not too shabby.  It's way faster than any of the Magneto-Optical formats. Although it's not stated anywhere, I'm assuming this is a random-access format, so it's access time during restores or retrievals would be very nice when compared to tape.  Due to the load/unload process, it's still not going to be as fast as a hard drive unless we're talking about leaving the disc in the drive all the time.  In a robotic setup, you'd have to add robotic time and load/unload time.  But this would all be similar to, if not better than, the speeds we have with tape.

The next question is cost, and there's nothing on that yet.  Traditionally, other optical formats have lost this race in a big way.  Only time will tell whether or not this format will change that pattern.

Finally, there's the question of long-term stability of the media itself.  I previously posted about the differences of tape vs disk in this area, and how tape is actually more stable for longer periods of time than disk is.  However, this is holographic storage and I honestly have no idea what the long term viability of data stored on such a medium would be.   I'm leaning towards the idea that it would actually be very stable, but I know that other optical formats are not as stable as one might think they would be, so…  Only time and more research will answer that question, too.

Assuming that they address the cost concerns and my hunches are right about its long term stability, I'm really leaning towards this as a long-term archival medium — as opposed to a backup and recovery medium.  While 54 MB/s may sound like a lot, it's just not enough for today's large data centers.  Throughput doesn't matter much in archival situations, but random access does, making this really well suited to archive.

For those of you ready to dump tape or disk for anything that gives you the portability and cost of tape with the random-access nature of disk, it looks like you're going to have to wait a bit.


Written by W. Curtis Preston (@wcpreston), four-time O'Reilly author, and host of The Backup Wrap-up podcast. I am now the Technology Evangelist at Sullivan Strickler, which helps companies manage their legacy data

1 comment
  • I could really see this coming in handy with some tier-5 archiving storage IF and only if the media itself has significantly longer stability than tape or disk media today. So when Joe Blow reads the file he’ll only read once every 3 years, the stub file will point to a “disk” of these holographic storage media. At that point, I would assume that the cost would be less than keeping it on actively powered disk and still provided better random i/o than a tape.