Judge Peter Kirwan, Santa Clara County Superior Court (Jason Doiy / The Recorder)
SAN JOSE — For Intel, it’s a case that just won’t die.
Intel Corp. has spent 10 years fighting an unfair-competition class action that accuses the tech giant of manipulating performance tests to raise the value of its Pentium 4 processor. Now, with a trial less than a month away, Intel’s last-ditch efforts for a favorable pretrial outcome appear to be falling flat.
Santa Clara County Superior Court Judge Peter Kirwan issued a tentative ruling Thursday denying Intel’s motion for summary judgment. Intel’s counsel had argued plaintiffs couldn’t prove Pentium 4 users paid more as a result of Intel’s performance reviews.
“The court finds there are triable issues as to whether a causal connection exists between Intel’s benchmarking practices and higher P4 prices,” Kirwan wrote.
Plaintiffs lawyers, led by Girard Gibbs partner Eric Gibbs, are seeking about $300 million on behalf of consumers nationwide who purchased computers equipped with the Pentium 4 processor for personal use in the two and a half years following its launch in November 2000.
In a recent court filing, plaintiffs lawyers chided Intel’s defense for its dogged, eleventh-hour campaign to unravel the case, which relies heavily on internal Intel presentation slides and emails, some of which instruct the recipients to “do not forward” and “delete after today’s meeting.”
“Intel has sought dismissal of the case at least six times; deposed plaintiffs, their families and coworkers, and over 100 class members; and petitioned the court of appeals and California Supreme Court to find plaintiffs’ case unsuitable for a class action trial,” the plaintiffs’ brief states. “Now with only a month until the trial begins, Intel asks the court to conclude that plaintiffs case is so lacking in merit that the long-awaited trial is not even necessary.”
Intel has turned to Munger Tolles & Olson partners Gregory Stone and Daniel Levin in Los Angeles, along with Miriam Kim in the firm’s San Francisco office.
The Munger Tolles team challenged Kirwin’s tentative ruling in court Friday. However, with a trial date set for May 5, the judge seemed unlikely to change his mind.
“Those arguments should be made at trial,” he said. “Ultimately, it’s going to be the plaintiffs’ burden to show the damages.”
The Pentium 4 microprocessor has long been obsolete, but plaintiffs argue it was a poor product from the time it hit the market. Class lawyers accuse Intel of engineering tests to boost its interindustry performance scores, when in reality, they contend, the new chip couldn’t compete with the Pentium 3, much less competitors’ products.
One email obtained by plaintiffs refers to a performance analysis conducted by computer maker Dell Inc. “[T]he results were dismal,” an Intel executive wrote in the May 2000 email. “I was hoping Dell made a mistake [with] the measurements, but it turns out that they are accurate.”
In response, Intel executives designed two new benchmark tests that would favorably rate the Pentium 4, plaintiffs allege. The company hid its involvement to make the tests seem unbiased, according to plaintiffs’ complaint.
“Between you and me, it is important that you understand that Intel has been the catalyst, the driving force, as well as the developer of this benchmark,” an Intel executive wrote in an August 2000 email. Another internal email states: “This benchmark is probably as close as we can get to a marketing tool dream, but we must proceed cautiously.”
According to plaintiffs’ complaint, the company also adopted a strategy known as the “dirty dozen” program, distributing a dozen Intel-written applications that would give Pentium 4 high ratings when used in existing benchmark tests. One commonly used benchmark scored Pentium 4′s performance as 10 percent lower than Advanced Micro Device’s rival processor before the new applications, but 13 percent higher after, plaintiffs argue.
“Few more of these and we will be in cruise control,” an Intel executive wrote in July 2001, according to court records.
The Munger Tolles attorneys accuse plaintiffs of resting their case on a handful of emails written in “blunt or hyperbolic” language. They insist plaintiffs have no hard evidence to suggest consumers were influenced by performance test results or that those results drove up prices for end products containing Pentium 4 processors.
“No matter how provocative the tone of some of these documents may be, such evidence is immaterial to the core question presented by this motion: whether plaintiffs can meet their burden of proving that the alleged unfair conduct actually caused any cognizable consumer harm,” Intel’s motion states. “They cannot.”
Girard Gibbs partner Geoffrey Munroe says his team has uncovered additional documents showing Intel’s use of misleading benchmarks extends beyond the Pentium 4. “It’s something they’ve continued to do and the group responsible for it has expanded,” Munroe said.
Plaintiffs seek an injunction against Intel and compensation under California’s Unfair Competition Law.
Defense counsel referred questions to an Intel spokesman, who declined comment.
A LEGAL ODYSSEY
Skold v. Intel, 05-39231, initially filed in April 2004, has slogged through the courtrooms of five judges. Three attempts to certify a class of Pentium 4 users met rejection before the Sixth District Court of Appeal demanded another look in 2011. Former Santa Clara County Superior Court Judge James Kleinberg finally certified a class in 2012 covering residents of all states except Illinois.
At that juncture, Intel replaced its defense lawyers from Skadden, Arps, Slate, Meagher & Flom with the Munger Tolles team, who are now making their own stab at dismantling the class action.
Last week, Intel followed up its bid for summary judgment with a motion for class decertification, insisting that after years of litigation plaintiffs still can’t make their case that the company’s alleged manipulation of benchmarks raised the costs of computers.
“[M]ost of the class never saw or heard of benchmarks,” the motion states. “They ask the court to ignore this glaring deficiency and assume that the challenged benchmarks must have misled some people—exactly whom they cannot say—and that these unidentified people in turn misled consumers.”
The defense attacks statistical models prepared by plaintiffs’ expert, economist Russel Lamb, saying they cannot show whether or how much consumers overpaid for computers. Any increase in the price of Pentium 4 processors likely would have been absorbed by manufacturers, the defense claims.
Lamb has estimated Intel’s illicit gains at $216 million to $318 million. However, Munroe acknowledged there is uncertainty as to the direct impact on customers’ wallets. “[Lamb] knows the amount consumers paid is higher,” Munroe said at Friday’s hearing. “He doesn’t know how much higher.”
That could be an obstacle for plaintiffs, said Venable partner Angel Garganta, a class action defense lawyer who is not involved in the Pentium 4 case, though his firm represents Intel in other matters.
“It doesn’t sound like it’s possible to demonstrate damages on a classwide basis,” he said, pointing to the U.S. Supreme Court’s 2013 ruling in Comcast v. Behrend, which found that an economic model designed to measure overpayment could not support class certification unless it measured damages specifically resulting from the alleged injury.
It’s a case that Munger Tolles also cites in its motion to decertify the class of computer purchasers. For his part, Munroe is looking forward to finally arguing that issue in court, after a decade of what he described as “motion upon motion.”
“I was maybe fresh out of law school when it was filed, but I’ve been working on the case for seven years,” Munroe said, “and never imagined it would last this long.”
Contact the reporter at firstname.lastname@example.org.