Is NBT the Inktomi of Storage?

Talk about gluttons for punishment. Seven years and one Internet bubble later, two former top executives of Inktomi Corp. (Nasdaq: INKT) are back -- only this time, their eyes are trained on the storage market.

Back in the halcyon days of 1996, Jerry M. Kennelly and Dr. Steve McCanne spotted the need for Web caching to accelerate performance of the Internet, so they made this Inktomi's objective. Yahoo (Nasdaq: YHOO) bought Inktomi just last month, in a deal worth $235 million (see Yahoo to Buy Inktomi).

Now they want to do the same for file acceleration in the corporate world. Kennelly and McCanne's new company, NBT Technology Inc., has closed $6.6 million in a first round of funding, five months after it was founded. The San Francisco-based company currently has 12 employees (see NBT Technology Raises $6.6M).

Accel Partners and Lightspeed Venture Partners invested in the startup, with Jim Swartz of Accel and Christopher Schaepe of Lightspeed taking seats on the company's board. [Disclosure: Lightspeed Venture Partners is an investor in Light Reading Inc., publisher of Byte and Switch.]

So who are Kennelly and McCanne, and what are they up to?Kennelly, who is NBT's president and CEO, was previously EVP and CFO of Inktomi. Prior to Inktomi, he served in executive positions at Sybase Corp. (Nasdaq: SYBS), Oracle Corp. (Nasdaq: ORCL), and Tandem Computers.

The illustrious McCanne, currently NBT's CTO, is a former professor of computer science at the University of California at Berkeley. He co-founded FastForward Networks, a startup developing Internet broadcasting technology, which he sold to Inktomi for $1.3 billion in September 2000. [Ed. note: That's in Internet Funny Money Dollars, but still!] In light of recent events, we felt it would be prudent to double-check McCanne's credentials -- and it turns out he really was a professor at Berkeley. [Ed. note: Anyone heard from Ken Lonchar lately? See Veritas Fires Veteran CFO.]

McCanne is so hot, the MIT Technology Review named him one of its "Top 100 innovators" in June 2002, and a UK-based snack company even named their oven chips after him! McCanne's work on Internet protocols, packet filtering, network multicasting, and streaming media during his tenure at Lawrence Berkeley National Laboratories in the 1990's apparently earned him all this street cred.

Could this be why Accel and Lightspeed invested in the company? "What other reason is there?" asks Jim Swartz, partner at Accel. "Steve McCanne's a rockstar."

The funding certainly wasn't for its products, as NBT won't be shipping anything until at least early 2004. And its ideas, from what we can tell, aren't exactly original either.McCanne says NBT is solving a complex problem that can be tackled in many different ways. "We've come up with a set of innovations that solve the general case of access to data, which includes very challenging problems in areas like remote backup over wide-area networks and moving very large files over long distances," he says.

It's the same set of problems that at least four other startups -- including Actona Technologies Inc., DiskSites Inc., Storigen Systems Inc., and Tacit Networks -- are aiming to tackle. The supposed breakthrough from these companies is the ability to write as well as read to remote file servers, with as little delay as performing the same function on the LAN. In other words, there's no more storage in remote offices: It's all centralized, and files are shared over the WAN and potentially cached, depending on how often they are used.

Alan Saldich, NBT's director of product marketing, claims the company's target application is broader than just file caching, and it's not just a distributed file system either. "It's a simple architecture, broadly applicable to any client-server interaction," he says. [Ed. note: A Swiss Army knife approach, then. Perhaps it's a VCR and a microwave and a storage appliance all in one!] NBT says it has filed for three patents on its technology so far.

"There is huge expense spent on storage around the world because WAN performance is so poor, you have to replicate everything to get the performance," says Saldich. That may be true, but WAN-distributed storage is such a new market, none of the industry analyst firms has any projections for it yet. And although these other startups might be shipping products, not one of them has announced a customer yet.

Still, the others have a healthy head start over NBT, if the market materializes. "The opportunity to develop channels for this technology is now, as the tech market comes back," says Tim Williams, founder and CEO of Tacit (see Tacit Makes Funding Explicit