How the ‘ghost riders’ theory won Rambus trial

November 18, 2011

At the beginning of his closing argument way back on Sept. 20, Micron counsel William Price of Quinn Emanuel Urquhart & Sullivan told jurors that his client and its co-defendant, Hynix, had fixed prices on some computer memory chips. They did it 11 years ago, he said, when the tech bubble burst and memory chip prices were plummeting. “That activity was wrong, and there were victims,” Price said, according to this transcript. “But Rambus wasn’t a victim …. And, so, what Rambus has done here is they’ve taken something that we have told you from the beginning was true and said they were victims when they weren’t.”

Price told jurors about an instruction New York City used to give its bus drivers:  if there’s a crash, lock the doors. “The reason wasn’t to keep people from leaving,” Price said. “The reason was to keep people from jumping on and saying they had hurt themselves, [that] they had hurt their back. Those people were called ‘ghost riders.'” Rambus, Price said, was a ghost rider. It claimed to be a victim of Hynix and Micron’s wrongdoing but wasn’t even on the bus. Rambus and its lawyers at Munger, Tolles & Olson asserted that Micron and Hynix conspired to wreck the market for its proprietary computer memory chip; Price and Hynix’s lead counsel, Kenneth Nissly of O’Melveny & Myers, had to show jurors that their clients’ admitted price-fixing involved the market for an alternative memory chip; the failure of Rambus’s chip, they argued, was due entirely to problems with the product.

“The challenge was to say, ‘Yes, there was price-fixing, but not this price-fixing,'” Price told me.

Quinn Emanuel and O’Melveny met the challenge, though they had to wait almost two months to find out they’d won the case. Twelve jurors began deliberations in late September. On Wednesday they finally returned a verdict, concluding that Rambus had not proved Micron and Hynix engaged in a price-fixing conspiracy to keep Rambus’s chip from becoming the industry standard. (Here’s the verdict sheet.) The jury also found that Micron and Hynix didn’t conspire to disrupt Rambus’s relationship with Intel. Those were the only questions jurors had to answer, since everything else in Rambus’s case depended on jurors finding a conspiracy to harm the company.

Price said the trial was truly a bet-the-company case for Micron. In its closing arguments, Rambus asked jurors to award about $4 billion in damages. With the trebling permitted under antitrust laws, the potential $12 billion exposure exceeds Micron’s market share. (Rambus had everything riding on the outcome as well; as Dan Levine reported Wednesday for Reuters, the company’s share price fell 61 percent after the verdict was announced.)

The relationship between Rambus and Intel, according to both Price and Nissly, was one of the keys to the trial. In the late 1990s, Intel agreed to use Rambus’s proprietary computer memory chip in its microprocessors, rather than the industry standard memory chips. (All of the random access memory chips have acronym names, but I won’t gum up the story by using them.) The deal had the potential to make Rambus the king of the memory chip business, but it didn’t work out that way; the relationship between Rambus and Intel soured a couple years before the contract ended in 2002. Rambus asserted that Hynix, Micron, and co-conspirators including Samsung poisoned its dealings with Intel. Price, Nissly and their teams countered that Intel was dissatisfied with Rambus’s product and its work.

Rambus, intriguingly, called no live witnesses from Intel. That left an opening for Hynix and Micron, and they took it. “We believed their testimony was helpful,” Nissly told me. A couple former Intel execs testified, but the most dramatic testimony, according to Price, came from an Intel engineer named Paul Fahey, who was subpoenaed by the defense. According to Price’s closing argument, Fahey said that Intel was working all-out to adapt the Rambus chips for its processors, to the extent that Intel employees canceled vacations, worked weekends, and risked their marriages. Rambus, meanwhile, seemed to take a much more cavalier approach — which was explained, according to Price’s account of Fahey’s testimony, only when Rambus filed a patent suit against Hitachi, claiming that Rambus patents were infringed in the industry-standard memory chips its rivals were producing. Intel then realized, according to Price, that Rambus had a backup plan if its propriety chips didn’t supplant the chips other companies were selling: it would sue those other companies for IP violations. (That’s the so-called ‘Plan B’ that allegedly resulted in Rambus shredding documents in company-sanctioned ‘shred days.’)

“Fahey exploded when he testified about that,” said Price, one of Quinn Emanuel’s most experienced trial lawyers. “It was the most emotional testimony I’ve ever seen at trial.”

When Hynix and Micron offered jurors an alternative theory for the demise of the Rambus/Intel partnership, their defenses were aligned. The two defendants had slightly different explanations, however, of their involvement with Rambus’s propriety chip technology — and with Samsung, which was banking on the success of Rambus’s chip. Hynix lagged far behind Samsung in gearing up to produce chips using the Rambus technology, as opposed to the industry-standard tech. Micron, on the other hand, had no plans to adapt to the expensive Rambus chips. Rambus asserted that even though the two companies were differently positioned, Hynix and Micron worked with Samsung to keep prices high for chips with the Rambus technology so it wouldn’t proliferate, and the industry-standard technology would prevail.

The Samsung link worked to Rambus’s benefit, since Samsung witnesses testified, in playbacks of videotaped deposition, to fixing prices. (Samsung paid Rambus $900 million to settle both IP and antitrust litigation in January 2010.) But according to Price, Micron and Hynix were able to show that the Samsung price-fixing testimony didn’t add up to a conspiracy with them. Their arguments, he said, were actually stronger because they were positioned differently in the market for Rambus chip technology.

“We had different stories, and that really helps in an antitrust conspiracy case,” Price told me.

The trial went much more quickly than San Francisco Country Superior Court Judge James McBride had originally anticipated. Jurors had been told that testimony might go on until Thanksgiving, so the late September closing arguments were a full two months ahead of the original timetable. Both Nissly and Price told me they were optimistic as jurors began deliberations.

Then came the long wait for a verdict. Price said all three of the law firms had betting pools on how quickly jurors would decide the case. None of them expected deliberations to go on for longer than two weeks. Instead, for almost two months, all of the firms sent a lawyer to the San Francisco county courtroom to wait for a verdict. Nissly and Price didn’t go every day; Price actually tried a case at the U.S. International Trade Commission — against Rambus — while deliberations wore on. O’Melveny partners Susan van Keulen and Tad Allan spelled Nissly and Quinn partners Jon Steiger and Patrick Shields manned the courthouse for Micron.

When word came Wednesday that the jury was back, Nissly made it to court for the reading of the verdict. Price was in Los Angeles, and sat by a computer waiting for news of the result. (Oddly, he first found out the jury had cleared his client from a Rambus shareholders’ blog.) Both lawyers said it’s still a mystery why the jury — which declined to be interviewed by lawyers immediately after the verdict — took so long to decide the case. Price said they didn’t ask the court many questions and never appeared to be deadlocked, but apparently went slowly through the evidence. “I’d still like to know why it took eight weeks,” Nissly said.

Rambus told Reuters last night that it’s reviewing the record for a possible appeal, though Price said “there isn’t much for them to complain about.” (As he pointed out, he’s a bit biased on this point.) Price said one argument Rambus might made — that the judge instructed jurors incorrectly on what standard to apply in deciding the conspiracy’s impact on customers — isn’t really relevant since jurors never reached the question of the alleged conspiracy’s impact.

For more of my posts, please go to Thomson Reuters News & Insight

Follow me on Twitter

No comments so far

We welcome comments that advance the story through relevant opinion, anecdotes, links and data. If you see a comment that you believe is irrelevant or inappropriate, you can flag it to our editors by using the report abuse links. Views expressed in the comments do not represent those of Reuters. For more information on our comment policy, see