We need a data storage museum

Core memory, Credit MIT archives, as displayed at the Museum of Computer History
Core memory, Credit MIT archives, as displayed at the Museum of Computer History
Just returned from Washington DC in the States. It has been decades since I had last seen the Smithsonian American History Museum and I was very saddened to see that their display of information technology of the 40s and 50s had been removed.

It’s unclear to me whether we have a museum of storage anyplace but I think we need one. For starters, technology is changing so rapidly that to see what we had even a few generations ago would take effort.

Why would someone want to see what was done 10 or even 50 years ago in storage. I can think of many reasons:

  • Innovate on prior technology – Innovation often takes a step backward before it can move forward. By displaying technology that existed 10, 20 to 40 years ago and explaining what it was made of, how it worked, and what it could do for industry someone might be able to innovate on these concepts into an entirely new storage direction.
  • Visibility of previous technology – The technology of miniaturization has left much of what’s interesting today almost impossible to see and touch. Yesterday’s technology was much easier to see, touch and almost walk through. As such, it was much easier to understand. Educating tomorrows engineers often starts with understanding yesterday technology.
  • Understand product lifecycles – Products go through a sometimes protracted technological evolution. Lifecycle changes can be supportive as well as destructive of product success. As such, showing how some products changed from generation to generation may help future engineers evolve better products.
  • Retrieve/recreate old artifacts – Recently, NASA had to retrieve old video feeds from their moon missions and had to revive some old technology to do this. Having a museum chartered with maintaining and displaying old storage technology might have made this job easier. And as more and more of the world’s information is retained as data, this problem only gets worse.
  • Diversify today’s technology foundations – storage technology has advanced so fast because of it’s singular dependance on electronic semiconductors to shrink and scale. Older technology was much more diverse using mechanics, magnetics, and electronics and as such, not so dependent on one technology. By having a place where the older technology could be researched, whole new breakthroughs, not dependant on today’s semiconductors could emerge that could take storage in totally different directions.

Probably many more reasons why a museum of storage would make sense but these will suffice. I have seen IBM, EMC and other companies product and technology historical displays, but they are not open to the public and there is no central repository for the storage industry as a whole. The problem with corporate displays of technology is that it only represents one view of technological evolution – the survivor’s view. Technological history must include more than just the victors, it must include the life and evolution of the defeated as well.

A storage museum could become a repository for any and every company’s storage technology. Today, for most companies, such technology is often scrapped or at best stored in a basement someplace. Such technology could be brought to light, researched and placed into perspective at a museum of storage.

I think we need such a museum before we lose the technology of the early years of data processing and fear that much of that technology is already gone, more disappearing every day. Also, the history of storage is written mainly in hardware evolution and as such, is easily displayed. As more of storage evolution is implemented via software functionality, it becomes less tangible and thus more easily discardable. If we start with storage hardware, maybe somebody will begin to appreciate the value in recording storage software history as well.

The American Smithsonian has a charter to study the history of technology. Some of this has been on display in Washington and perhaps, at the Museum of Computer History in Mountain View. But nowhere is there a focus on data storage and I believe there should be. How such a museum would be funded, where it would be situated, and how it could be started must be subjects of future posts. Comments?

What really drives storage innovation

Ongoing waves of consolidation remind me of what really drives storage innovation – companies willing to experiment. Startups can only succeed when their products can engage the marketplace.

Startups risk everything to develop technology an innovation or two that can change the world. But what they ultimately discover, what they truly need is some large and/or small company’s willingness to experiment with new and untried technology. Such market engagement is essential to understand their technology’s rough edges, customer requirements, and distribution options.

Recently, I was informed that some large companies prefer to work with startups because they can better control any emerging technology direction. Also, their problems are big enough that typically no one solution can solve them. Startups allow them to cobble together (multiple) solutions that ultimately can solve their problem.

From the small company’s perspective the question becomes how to attract and begin the dialogue with innovative customers willing to invest time and money in startups. But, the real problem is knowing enough about a customer’s environment to know if suitable prospects for their technology exist. Armed with this knowledge, targeted marketing approaches can be applied to ultimately get a hearing with the customer.

However, what’s missing is a forum for large and small companies to describe their environment and more importantly, their serious, chronic problems. Mostly, this has been done informally or on an ad hoc basis in the past, but some formality around this could really benefit storage innovation at least from startups.

I see many possibilities to solve this, ways that companies could provide information on their environment and identify problems needing solutions. Such possibilities include:

  • an electronic forum something like Innocentive.com where companies could post problems and solicit solutions
  • an award to solve a particularly pressing problem like Xprize.org where a group of companies, perhaps in one vertical combine together to offer a significant award to help solve a particular nasty storage/IT problem
  • an organization of sorts like SNIA end user council that could provide anonymous information on IT environments and problems needing solutions.
  • a Small Business Innovation Research-like (see SBIR.gov) that could provide a list of problems soliciting solutions

The problem with SNIA end user council and SBIR-like approaches is the lack of anonyminity, the problems with an Xprize-like award is the inability for any one organization to fund the award. All of which is why I prefer an innocentive.com-like approach, maybe better targeted to IT issues and less targeted on basic and materials science. Finally, perhaps another, unforeseen approach that might even work better – comments?

Why big storage vendors can’t be enticed to work on something like this is another conundrum and probably subject for a future post.