By STEVE ADAMS

The Patriot Ledger

Larry Genovesi faced two major obstacles while making presentations to two dozen venture capital firms and trying to raise $8 million for his new company, Ammasso Inc.

Venture firms were reluctant to invest in risky technology startups in the aftermath of the high-tech meltdown.

And Genovesi was asking them to roll the dice on an unproven technology that competes directly with InfiniBand, an approach to computer networking that many of the same firms had already backed.

“It was one of two responses,” Genovesi, 45, of Cohasset, said. “It was either, ‘You’re wrong, it’s going to be InfiniBand,’ or we’d get the sort of deer-in-the-headlights stare which means, ‘Oh my God, I’ve invested a lot of money in a new technology that’s not going to work out.’ “

Fortunately for Genovesi, one of New England’s most active venture capital firms hadn’t jumped on InfiniBand’s wagon as the preferred networking option to connect clusters of high-powered computers. Westwood-based Prism Venture Partners led a syndicate of three venture firms that awarded Ammasso with $10 million in early- stage financing June 17.

Genovesi hopes to capture a share of the $300-million-a-year market for clustered computer networks. Within a few years, he estimates it has the potential to grow to a $1 billion market.

The potential clients: online merchants, research labs and biotech companies that need vast amounts of networked capacity to allow users to work simultaneously on complex applications involving trillions of calculations per second.

Genovesi, who founded Canton server manufacturer Network Engines in 1997, struck out on his own in 2002 after turning over day-to- day operation of the company to John Curtis.

Genovesi began talking with Bill Elliott, Network Engines’ vice president of marketing, about starting a new company in early 2002. Joined by another Network Engines executive, Thomas Tucker, they devised a business plan in the fall of 2002 and started making their pitch to about 25 venture capital firms.

The response was chilly.

“Although the amount of money they had available to invest was the highest it had ever been, they weren’t doing deals,” Genovesi said. “So it was very difficult. They were looking for every conceivable excuse to say no.”

Genovesi is convinced that Ammasso can design Ethernet-based networks that would provide unprecedented speed and power at a price 50 to 80 percent less than the Infiniband model. “My experience is Ethernet always adapts and overcomes,” Genovesi said. “Everybody’s always predicted the demise of Ethernet and it just doesn’t happen.”

InfiniBand has a head start. Developed in 2000 by a consortium of the largest technology companies, InfiniBand has been touted as the next-generation solution to linking large numbers of computers on shared applications. Its specifications promise data transfer at a speed of up to 2.5 gigabytes per second – which is more than twice as fast as the current Ethernet standard.

The challenge, Genovesi said, is to develop an application that solves two of the main limiting factors in networking: the strain on the main processor caused by transferring data to individual terminals, and latency, the delay when a message is delivered from one segment of the network to the next. If Ammasso’s final product can do that effectively, it will be on par with InfiniBand’s performance at a fraction of the cost, he said.

Since its protocols were adopted in 200, more than 70 companies have jumped into the InfiniBand market, with the first products expected to be shipped later this year.

In the past year, however, some of the major players in the technology sector have shown signs of dissatisfaction with InfiniBand’s potential. Intel announced last fall that it was developing an Ethernet chip that could trump InfiniBand as the application of choice for large cluster networks.

“When you get large customers like Intel pulling (out), you know that it’s in trouble,” said David Baum, a general partner for Prism who negotiated the deal with Ammasso.

Recent publicity about waning confidence in InfiniBand boosted interest in the Ammasso deal, Baum said. Through a syndication with CDP Capital-Technology Ventures and Inflection Point Ventures, Prism was able to raise $10 million in startup financing for Ammasso. Most of the money will be spent on engineering.

During five months of negotiations with Prism, Genovesi was asked to provide detailed business strategies and product plans that he normally wouldn’t be asked to furnish until after the company was up and running. Such are the tradeoffs when cutting a venture capital deal.

“They really put us through the wringer and that’s a good process, but when you’re waiting to get funding it can be painful,” he said.

This week, Genovesi could look back at the experience with some distance. Ammasso’s 20 employees are settling into their new quarters in South Boston’s Fort Point Channel area, where Ammasso is renting 19,000 square feet on two floors in an industrial loft. Genovesi expects to hire 10 more by the end of the year, with the hope of bringing a product to market by next spring.

Ammasso, which is Italian for “cluster,” fits in well in this scattering of former factories and warehouses now being colonized as artists’ lofts, galleries and tech startups. It’s the sort of laid- back workplace where some employees use their first names as their E- mail addresses. Genovesi’s poodle Sparky takes advantage of Ammasso’s pet-friendly environment.

Ammasso’s strategy is to work with a small number of business partners to handle the sales and distribution side of the business, with the goal of signing contracts with major server manufacturers like Dell and IBM.

In an era where businesses remain vigilant about capital spending, Ammasso’s products will allow them to expand their data storage and processing capabilities gradually.

“Instead of having to buy very expensive, high-end mainframe boxes from server vendors, you buy inexpensive commodity servers, hook them together in a cluster and run your software in a cluster with very good performance and significantly less money,” he said.

And Genovesi is convinced that he’s received funding at just the right time.

“If you believe the economy’s going to turn around the next 12 to 18 months, now is the time to invest in technology startups,” he said. “They’ll have the product when the economy is growing.”

Steve Adams may be reached at sadams@ledger.com.