|Hynix Lawyer Says Rambus ‘Squandered' Memory Chip Opportunity |
Sept. 20 (Bloomberg) -- A lawyer for Hynix Semiconductor Inc. told a jury that Rambus Inc.'s flawed technology and poor business decisions, rather than collusion by rivals, hobbled its chances of making its memory chip the industry standard.
In his closing argument yesterday at a state court trial in San Francisco that began in June, Hynix lawyer Ken Nissly denied that the world's second-largest maker of computer memory conspired to drive chips designed by Rambus out of the market.
Nissly showed jurors a timeline titled “Rambus Choices Doomed Intel Relationship” to introduce his argument that Rambus “had the best support for innovation they could possibly have” in its 1996 contract to develop its own version of dynamic random access memory, or DRAM, with Santa Clara, California-based Intel Corp., the world's largest chipmaker.
Rambus “squandered that advantage such that Intel decided by the end of 1999 that it had made a mistake, that it was going to change its road map,” Nissly said.
Rambus contends that Ichon, South Korea-based Hynix and Boise, Idaho-based Micron Technology Inc. colluded to lower the prices of their own memory chips and deserted their commitment to produce Rambus-designed DRAM, or RDRAM, relegating it to a niche role. Micron's lawyer is scheduled to give a closing argument today after Nissly finishes.
Rambus, which doesn't manufacture the chips it designs, argued it would have earned $3.95 billion in royalties without the alleged conspiracy. Under California law, a jury finding of damages in that amount would be automatically tripled to $11.9 billion.
‘Rigged the Race'
Hynix and Micron “cheated” and “rigged the race,” Sean Eskovitz, a lawyer for Rambus, said yesterday in his closing argument. The two chip manufacturers first fixed their prices below market value and “when they thought they had pushed RDRAM off to the side, they jacked up the prices.”
“Why did they do this? Power and money,” Eskovitz said. “They knew RDRAM was a threat. They knew they needed to shoot it over and over and over again,” and “they used every possible tactic available to them.”
The chip manufacturers, in denying the claims, are relying in part on testimony from managers at Intel, which in the 1990s selected RDRAM as a solution to a computer-memory bottleneck. Intel managers said the collaboration ultimately failed because RDRAM was flawed and because of a contract provision potentially allowing Rambus to block Intel from shipping processors relying on the chip designer's technology if certain conditions requiring Intel to promote RDRAM weren't met.
Eskovitz showed jurors a chart depicting the price of RDRAM dropping in the summer of 2000, after its introduction a year earlier. He highlighted video testimony and e-mails between executives and sales people at Micron and Hynix that he said demonstrate how the chip manufacturers intensified their collusion to stall production of RDRAM.
The collusion drove RDRAM prices high enough to cause Intel to alter its memory “road map” to include Double Data Rate chips made by Hynix and Micron, and blocked RDRAM from becoming an industry standard, Eskovitz said.
“Intel wanted RDRAM to succeed, Intel picked RDRAM because it was the superior technology,” Eskovitz told jurors. William Swope, a former Intel manager, testified “Intel's job was to have superior performance,” Eskovitz said.
Rambus played video-taped testimony from Mike Sadler, vice president of corporate development at Micron, who said he ceased e-mail communications about chip pricing in 2002 because the company received a U.S. Department of Justice subpoena as part of a price-fixing investigation.
Eskovitz reminded jurors of Judge James McBride's instructions that they may consider the fact that Hynix, in 2005, and some of its employees, pleaded guilty to U.S. charges of fixing prices of some chips. Micron avoided prosecution in the price-fixing probe by cooperating with the government.
Nissly began his closing arguments by claiming that Eskovitz gave jurors dates and events “wadded up and thrown together.”
He said Eskovitz avoided addressing the central question: Did Micron or Hynix conspire to fix RDRAM prices high and DDR prices low, or restrict RDRAM output in order to prevent RDRAM from becoming the standard for computer memory?
The case is Rambus Inc. v. Micron Technology Inc., 04-0431105, California Superior Court (San Francisco).
--Editors: Peter Blumberg, Mary Romano