Holographic Storage And Other Science Projects

There's been a lot of news over the past month or so about the fringe data storage technologies I think of as science projects. You know, holographic memory, 35TB tapes, 10 layer 200GB Blu-Ray disks, optical tape, isolinear data rods and data crystals.  Oh wait -- the last two are science fiction, for now. The biggest news was the Colorado Department of Revenue seizing the assets of InPhase Technologies after it was shuttered last week.  InPhase, a 10-year-old Bell Labs spin-off, has been demonstrating a 300GB holographic disk and promising to deliver in 6-12 months for at least the past three years.

Because holographic storage, in one form or another, is the data storage and transfer medium of choice in most science fiction, most geeks have been hoping to see a commercial holographic system. Unfortunately InPhase's delays allowed deduplication, CAS and other technologies to mature to the point where Tapestry's 300GB capacity and  20MB/s transfer rate couldn't justify the $10,000+ price of a drive. I'm sure someone else, maybe GE, will try holographic storage again in the next few years. It's just so cool. On the other hand when he heard of InPhase's demise David Vallente of Wikibon tweeted a quote from storage industry eminence grise, Fred Moore: "Holographic Storage is the technology of the future - and it always will be."

Then there was the IBM/Fujifilm 35TB tape story. This seems like a near-future science fiction story, like Arthur C. Clarke in 1940s, predicting satellites, not warp drive or transporters, I'm sure I'll see a 35TB tape in my lifetime. I'm also sure I'll retire any hardware I'm buying today before I see one (and I keep my hardware a long time). I still have a Dell PowerEdge 1550.

Finally, there is David Hill's blog entry on Opternity's laser tape promising ten times the data storage as current magnetic tapes.  I think they probably have a market if they can ship today.  Three years from now it they'll be trying to sell a whole new technology for 10GB cartridges when LTO-6 can hold 3TB. Looking out seven years or so, those 35TB tapes from IBM/Fujifilm might come to market at the same time. This illustrates one of the defining features of a science project: it looks great compared to today's alternatives, but no one bothers to compare it to the conventional technology's roadmap.

I call another common feature of science projects the "everything old is new again rule."  In the case of Opternity, they reminded me of a similar product from 1989.  ICI, a British chemical company and Iomega announced "Digital Paper," a thin film optical medium that could be used as tape for sequential access or in one of Iomega's Bernoulli cartridges for random I/O.  Of course, it never made it to market. All in all, you have to love the science projects.  After all, bubble memory, NAND flash and even RAID were once science projects too.