There are two worlds in storage and suppliers generally fit in one or the other. The large established players have invented and developed their starting offering and then internal product creativity withers. With size and success comes stifling caution and creeping fear of cannibalising their own products.

The startups invent, think 'out of the box', 'off the wall', and bring radical new products to market. Some succeed, many fail. The successful ones either grow big, like Mr. Dyson and his bagless vacuum cleaners, or else get bought and acquire trousers with deep, deep pockets for the millions of dollars their founders earn.

This generalisation is broad and crude. Not all large successful suppliers are non-inventive. But they all generally have trouble productising their inventions.

Think HP's storage grid. Certainly HP has invented. But all it's got to show for its smart cell-based grid is RISS, a Centera-like competitor. It's a good product but what we don't see are smart-cell based SANs, NAS grids or WAFS capabilities or anything else. Smart cells are too darn smart for their own good, or the fundamental design, of having storage and processors allocated to each other in a product is wrong, or something.... Whatever, HP's smart cells are in danger of being left behind.

Take IBM storage research efforts - they're wonderful, well in front of the leading edge, but they never seem to actually become leading edge. Remember the atomic scale hammers? Remember holographic storage? It isn't IBM that's productising holo, it's InPhase.

Sun's R&D efforts are great but, again, we wait, and we wait, and we wait for stuff to hit the market, to excite us. CEO Jonathan Schwartz mentioned another hot storage R&D project at Sun in a Gartner symposium day recently: " He gave an example of a quad-core Opteron system with 24 terabytes of storage running Solaris and the ZFS files system as a product of R&D that has economics that 'no one can match.'"

No doubt its R&D prospects are great, terrific. When is it going to hit the streets? Will it hit the streets?

The recent big storage inventions we know about all came from start-ups with the big storage vendors jumping on the new product band wagon through OEM/reseller deals or through acquisition. Let's run through a few:-

- Clustered NAS - Acopia, Exanet, Isilon with NetApp acquiring Spinnaker, Brocade acquiring NuView and Cisco building a funding relationship with NearPath
- DPM - Crosswalk and Bocada
- Continuous Data Protection - Revivio.

The big exception to this? EMC's Centera. That was a game-changing storage development. HDS' TagmaStore development was also an eye-opener. It certainly sets HDS apart as an array vendor but no other array vendor has followed suit. These apart, what other storage development from a big vendor has set the market alight recently? I can't think of one.

If a big vendor could combine its market reach and product reliability with the creativity and energy of a successful storage start-up then what a company that would be.