Storage Outlook: What's on the Storage Horizon? Page 2
What's on the Near-Term Horizon (12-24 Months)
Almost every storage pundit on the planet says we are just on the horizon for optical storage that will solve the world’s problems. The problem is that this has been said off and on for a number of years. Perhaps this time it's true, but I do not see it happening within two years for the following reasons:
- The cost for optical storage will be high initially, which limits its usage
- If the cost is high, the only place it might be used is where organizations can afford the cost (large enterprises and/or the U.S. Government), which means it must be proven technology with a long shelf life and low bit error rates
This might be a boring prognostication, but what I see coming in the near-term is more of the same. I expect tape density to increase with LTO and enterprise tapes. In fact, if you look at the trend in tape density over the last few years, it is increasing at a faster rate than disk density (60 GB native for StorageTek 9940 tapes in 2000, and 200 GB native in 2002). This trend will continue given that some significant technology improvements have been made in the development of high-density tape and tape drives.
What's also significant is the actual amount of data written on the tape with compression. The trend from some of the HPC sites I work with is significant. In one example the site was seeing 1.3 to 1 compression using the 60 GB tape drive. After moving the same data to the 200 GB tape drive, they saw 1.6 to 1 compression. That means the capacity of the tape increase was not 140 GB but rather 242 GB, or 4x over two years. Of course there are many applications that cannot be compressed, such as movies, audio, some pictures, and others, but a great deal of data on tape is compressible.
What is on the near-term horizon from a bunch of different companies is power management ATA storage. The idea is that disks running without power can be easily powered on when someone needs to write/read data. Serial ATA (SATA) drives have a power management interface that allows this, and laptop drives have had this power management interface for years.
What's on the Mid-Term Horizon (24-36 Months)
This is a hard area to predict for several reasons:
- Standards groups do not always follow schedules — I know of a number of products that were delayed because of the change from 1-Gig FC to 2-Gig FC. Whatever you say about a new storage technology, the interface will have to be something that follows the standards. Of course, there will be market leaders and followers, but still, this is not the 1970s.
- What are going to be the tradeoffs between density and speed? Will two technologies develop that will allow each, or will one provide both? Some optical technologies are expected to provide both. Something that does not have high performance is going to be an issue given the time to migrate the data to the next product.