Computer Firm Lays Off Nearly Half Its Staff : Cutbacks: Move is in response to Supercomputing Solutions’ second-quarter loss of more than $1 million.

Supercomputing Solutions, a San Diego-based supercomputer manufacturer, announced Wednesday that it has laid off nearly half its 60-member staff, the first of several cost-saving measures designed to remedy the company’s financial woes.

The 4-year-old company dismissed about 30 employees--a mixture of administrative, production and technical staff--at the close of business Tuesday, said Jerry Canning, the company’s chief financial officer.

The cost-saving measures, including cutbacks in travel and capital expenditures, were announced after release of the company’s second-quarter and year-to-date earnings reports, which showed significant losses.

For the second quarter ended June 30, the company posted a loss of $1,113,756 on revenue of $738,723, contrasted with a loss of $743,836 on revenue of $360,752 for the same three months last year.

Canning said the company’s financial woes are due to high research and development costs related to the company’s primary product, a supercomputer called the CAPPS-9064. The company so far has been able to sell only one of its machines--to TRW--since putting the product on the market in June. The computers cost from $400,000 to $1.8 million each, depending on added features.

Supercomputing Solutions, which leases a plant in Sorrento Mesa, markets its supercomputers and software to the aerospace, defense and electronics industries as well as to U.S. government agencies.

But attracting new customers is a difficult task in the extremely competitive $1.5-billion supercomputer market, said Christopher Willard, a senior industry analyst at Dataquest Inc., a San Jose-based high-technology market research firm.

“A lot of companies go out of business right around this point,” Willard said, referring to Supercomputing Solutions’ heavy R&D; costs. “The company spends all of its money to get its product out the door. Then they don’t have enough cash to hold out until the orders start coming in. In this extremely competitive market, companies end up starving to death.”

Purchases of supercomputers have slowed, in part, because “peace has broken out,” Willard said. “Companies that do business with the Defense Department won’t be looking to buy new supercomputers.”

Supercomputers are considered the computer industry’s most powerful machines and are measured in terms of processing speed, accuracy and memory capacity. They can complete difficult computing tasks that would be impractical or impossible on less powerful computers.

For example, national laboratories, universities and the Defense Department use supercomputers, costing as much as $25 million, for such diverse purposes as simulatiing models determining the chances of galaxies colliding, chemistry computations and economic forecasts, Willard said.

Such major supercomputer users have already invested in these machines and are unlikely to buy new models, he said. As a result, manufacturers, including Supercomputing Solutions, are increasingly designing cheaper products for industrial users.

“For example, private industry can use supercomputers for product development,” Willard said. “Automakers can either build cars and drive them into walls, or simulate such crashes on the computer.” Supercomputers for private industry cost from $400,000 to $2 million.

“But now everybody’s rushing to supply industrial users, so competition in the low-end market has intensified, too,” Willard said.

Supercomputing Solutions says it has enhanced its machine by adding “parallel processing,” which allows dozens of powerful computers to be linked to tackle mammoth assignments.

“In theory, parallel processing sounds good,” Willard said. “But developing software to make all these machines run together and work together on a single problem has been a major” hurdle.

“People have been working on parallel processing for the last 15 years,” Willard said. “It takes a real genius to make it work.”

Supercomputer Solutions, founded in 1986, was first known as General Microelectronics Corp. In March, 1989, General Microelectronics formed a joint venture with Tinton Falls, N.J.-based Concurrent Computer to establish the subsidiary Supercomputing Solutions. Four months later, General Microelectronics bought out Concurrent’s 50% interest in the subsidiary and replaced its name with Supercomputing Solutions.

For the six months ended June 30, the company’s loss stands at $2,505,968 on revenue of $1,129,816. These results compare to a 1989 first-half loss of $1,241,815 on revenue of $817,363.