It struck me looking at Storage Tank and HP-s RISS, EMC's InVista and Sun's X4500 aka Thumper, that storage giants are very inventive with new storage technologies. But they don't invent products that will eat their current product's lunch. That's left to competitors to try.

So we have the majors espousing various forms of utility computing but it is 3Par that actually delivers an integrated storage product that can do it. Buy a 3Par array and you find, 3Par hopes, that you no longer need to buy drive arrays from EMC, or NetApp, for example.

The virtual tape technology we are surrounded with comes from FalconStor, Sepaton and others, not HP or IBM or StorageTek, at least not in the open systems world.

Major vendors invent storage technology to extend a product range or to defend it. Thus HDS invents disk array virtualisation in its TagmaStore controller and enables it to virtualise other vendors' arrays.

McData adds an intelligent platform to its director range because, if it didn't, Cisco and Brocade would, and are doing so. No director IQ means declining director sales and so the cycle of SAN fabric technology inventiveness goes around the block again.

Major vendors buy in inventive technology that renders existing product less effective or points out holes. Thus HP bought AppIQ for its SMI-S-based storage management software. It was quicker to buy it rather than make it, prevented a competitor getting AppIQ and filled a looming gap in its system management offerings.

Ditto NetApp with Alachritus and Spinnaker.

However the major has to be quick to implement otherwise a competitor flashes past. Such has been the case with Isilon and NetApp. Isilon had no installed base of products to integrate its new technology with. NetApp had to fir Spinnaker technology to its existing product range, which has taken a couple of years. That meant Isilon has been able to bring its very fast cluster-based NAS system to market faster than NetApp.

New and inventive storage companies now may intend to grow and become the new EMC, a great gorilla in a major storage niche, or they may be intent on being bought by their deep-pocketed elder brethren.

The major companies can offer storage inventiveness that reinforces the product infrastructure they have. Customers then hopefully will buy from an infrastructure vendor rather than getting best of breed for this, best of breed for that.

If the customer problems are acute enough then best of (new) breed wins and so companies like 3Par, Acopia, Isilon and others progress and grow. Many customers have milder problems and will wait for a major vendor to bring the technology they need to market.

Sun's VSM Open problems has come about because a make vs buy decision went for make and turned out to be wrong. No doubt Sun will now make a buy decision, or OEM, or resale, very quickly.

The major storage suppliers are generally very good about inventing their own technology or buying someone else's. A lot of venture capitalists and storage technology company founders are very much richer as a result.