To myself and a few of my friends, the chart at right brings back a lot of memories. We used to work for Stac (where the Stacker disk compression utility was developed). In the early '90s, we noted the steep decline in the cost per megabyte of disk storage, and wondered what that meant to the future of disk compression. We concluded that disk cost was trending toward free (quite rapidly!), and that when disks were free, the value of disk compression would approach zero. Then we thought about the implications of “free” disk space, and what sort of software might leverage it. We came up with the notion of backing up to disk (instead of the then-pervasive tape backup), and the idea of compressing the data from many computers through what is today called “de-duplication”. Unfortunately for us, and for many reasons, we never figured out how to turn that idea into a lucrative business.
Here's something that looks very much like an updated version of the chart we created back in the early '90s:
Matt Komorowski is the fellow who put this data together. He also calculated the regression equation (at right). He pointed me to another trove of data on historical disk storage costs.
Reading through all this stuff was a real trip down memory lane for me...
Stac Electronics stands out as one of the most innovative companies I've had the opportunity to work for. I was fortunate enough to work with some really brilliant developers, to learn a great deal about file systems, disk utilities and device drivers there and then later when they acquired Reachout, I had the opportunity to learn about various communications protocols among other things. So much of the code for both products was in assembly (including even the UI for Reachout) that it was a wonderful chance to learn how the hardware and the operating system works in much greater detail and we spent a lot of time with books like "DOS Internals" and "Undocumented Windows".
ReplyDelete