The networking giant announced its intention to buy NeoPath yesterday. It did not disclose financial details of the deal, which it said will close within the next four weeks.

Founded in 2002, NeoPath late last year was claiming just over 50 customers. The company has raised $29m venture capital, including a stake taken by Cisco in 2006, when Cisco was rumored to have negotiated first refusal rights for the acquisition of the start-up.

NeoPath sells systems that virtualize or aggregate multiple cross-vendor NAS or file-level storage devices into single namespaces. This is a small and developing sector with limited sales, in which NeoPath has only a handful of rivals. One of those competitors is EMC, courtesy of the NAS v-word systems it acquired when it bought Rainfinity Inc in 2005.

EMC is also one of the largest resellers of NetApp’s storage networking or SAN equipment. Not wanting to provoke any speculation that even limited NAS v-word competition could jeopardize its relationship with EMC, Cisco yesterday refused to say what its plans are for NeoPath’s products.

Instead Cisco said that by acquiring NeoPath it will gain the expertise that it says it needs to develop interfaces that will allow EMC and other vendors to run their file-level software on Cisco’s networking devices.

We had no expertise at the file system level and [with NeoPath] we’re getting people who do, said Cisco’s vice president of product management Ed Chapman. By getting these people in, we can figure out ways to siliconize file level opportunities and expose APIs to our partners, Chapman said.

That will not be immediate. There won’t be an ASIC tomorrow. But we’ll be able to offer our partners a range of APIs, he said.

Cisco has already developed smart SAN switches that host third-party block-level storage applications such as EMC’s Invista block-level virtualization tool.

So this would be much what we’ve done at the block level, but at the file level, Chapman said, referring to Cisco’s smart SAN switches, which can run third-party applications such as EMC’s Invista block-level virtualization tool.

But since their launch around three years ago, customers have been almost entirely indifferent to the third-party applications running on smart SAN switches. Granted, storage virtualization is not where people five years ago thought it would be now. But people are using other things that we’re providing, such as accelerated data replication, he said.

Cisco claims that until the acquisition deal has closed the company cannot declare its intentions for NeoPath’s systems on the grounds that if it did so, it would immediately affect NeoPath’s business. According to Chapman that would be the case even if Cisco were to guarantee the future of NeoPath’s systems.

If we said we were going to carry on selling them, customers might delay their purchases until after the acquisition, he claimed.

NeoPath’s systems would be far from the first storage products in Cisco’s portfolio. Cisco has been selling Fibre Channel storage networking switches since 2002, and has been offering wide-area file services systems originally developed by Actona Technologies Inc since it bought that company in 2004.

What about Actona’s expertise in file-level data processing? Certainly the Actona team understands how protocols such as CIFS and NFS work, no question. Neopath simply gets us expertise deeper into those technologies, Cisco told ComputerWire in an email.

NAS virtualization systems aggregate multiple heterogeneous NAS devices into single namespaces. While the nascent market for such systems is expected to grow, at present the technology is only developing, and its appeal is limited to big customers who want to consolidate large numbers of NAS filers behind a v-word system.

Alongside EMC, NeoPath’s rivals are fellow start-ups Attune Systems Inc and Acopia Networks Inc, and to a lesser extent Brocade Communications Systems Inc with the technology it gained when it bought NuView in 2005.