3Leaf ends up at Huawei – but will it be staying there?

Contact: John Abbott

Six months ago, I/O virtualization startup 3Leaf Systems disappeared from our radar screens. A little digging around more recently revealed that key staff members had scattered. VP of marketing Shahin Kahn was now at ORION Marketing Group, a consulting firm, with other ex-Sun Microsystems colleagues. CEO B.V. Jagadeesh had turned up as CEO of Virtela Technology Services, a managed network, security and technology services company. In his company biography he revealed that 3Leaf had been sold in a ‘private transaction.’ The trail of clues led on to Bob Quinn, founder and CTO of 3Leaf, who could be traced (via his LinkedIn profile) to Chinese telecom equipment maker Huawei Technologies, where he was now acting as a consultant. We surmised, and later received confirmation, that Huawei was the new owner of 3Leaf’s technology.

Two weeks ago, Huawei submitted an application to CFIUS – the US government’s Committee on Foreign Investment in the United States – including its first public statement that it had acquired 3Leaf in May. No details emerged other than that only the intellectual property and 15 of the 50 employees had been obtained in a transaction worth around $2m. According to CFIUS, Huawei should have requested permission from the committee. Huawei said it regarded the deal as a patent sale and hiring exercise and so believed it didn’t need to clear it with CFIUS. In 2008, the company abandoned its bid for 3Com due to US security terms. Hewlett-Packard stepped in to acquire 3Com a year later. More recently, Huawei has faced opposition to a proposed equipment-supply partnership in the US with Sprint Nextel over security concerns.

Aside from all this, the deal is a sorry – and somewhat worrying – end for 3Leaf, which raised roughly $67m from VCs Alloy Ventures, Enterprise Partners Venture Capital and Storm Ventures, as well as money from strategic investors Intel and LSI. 3Leaf was working on what should be a hot sector, I/O virtualization, but perhaps it entered the market too early. Its first product, the V-8000, first shipped in May 2007, but used a somewhat proprietary approach due to the lack of standards at the time. The company effectively started all over again in 2009 with plans to build new technology for virtualizing CPU and memory resources across x86 server clusters. It was looking for deals from OEMs, although there were also plans to sell prepackaged versions based on SuperMicro servers. However, 3Leaf needed more money to fund the ongoing research, and those efforts appear to have been unsuccessful.

3Leaf looked promising when it was founded, but early technology decisions led it down a blind alley. There may be some value in its patents and certainly more in the experience of its engineers, but it seems unlikely that, if CFIUS forces Huawei to sell the assets it’s bought, there will be many takers. If one is found, it’s likely to be a major server vendor with networking pretensions such as HP or IBM, or an I/O and networking adapter specialist such as Emulex or QLogic. Meanwhile, other startups in closely related areas – including ScaleMP, NextIO, Numascale, RNA Networks, VirtenSys and Xsigo Systems – soldier on.