Vol 14, Issue 17

April 24, 2000

Ethical Standards

By Peter N. Glaskowsky

Rambus and Hitachi are currently involved in legal proceedings that hinge on a single narrow question—did Hitachi violate Rambus's intellectual-property rights? This simple question, however, has implications that extend well beyond the two companies involved.

The outcome of these proceedings will be directly relevant to every company making or using synchronous DRAMs. Rambus's case against Hitachi (see MPR 2/7/00-04, "Intel, Rambus Draw Lines in Silicon"), recently taken to the International Trade Commission and expanded to include a total of eight patents, covers technology that is essential to all PC100, PC133, and DDR SDRAM designs. The Rambus patents are also likely to relate to the work of the Advanced DRAM Technology (ADT) effort, announced just one day before the Rambus lawsuit.

Hitachi has filed a formal answer to Rambus's allegations. In its response, Hitachi describes several "affirmative defenses," legal arguments it claims protect it against the charges. Of these arguments, the most substantial relates to Rambus's participation in the JEDEC committee that developed standards for SDRAM devices and modules. Hitachi says that JEDEC policies predating Rambus's participation require JEDEC members to disclose any patent applications that relate to the work of JEDEC committees to which these members belong.

Rambus, according to Hitachi, did not notify JEDEC that it had previously applied for patents covering specific elements of the SDRAM standard then under development. Rambus filed its original application in April 1990, more than a year before it began attending the relevant JEDEC committee meetings. The company disclosed one patent to JEDEC in 1993 (one that is not among those at issue today), and in 1995 it was careful to disclaim any position on the Synclink proposal, then under development in another JEDEC committee. These actions suggest Rambus was aware of the JEDEC disclosure policy.

Hitachi, in essence, argues that Rambus's silence on possible intellectual-property claims to SDRAM technology amounts to a free license to the IP that Rambus had developed but not disclosed. Hitachi does not make a particularly strong case for this position in its court filing, but I believe such a case could be made.

In my opinion, the critical question is this: Does participation in a standards committee create an implied contract among the members of the committee? Contracts require an exchange of value under agreed-upon terms. I believe this requirement was met in this case. The value lies in the cooperation among JEDEC members, and the terms of the agreement were expressed in JEDEC's policies.

Some standards organizations sidestep this whole question by requiring members to join a patent pool. To join the PCI Special Interest Group, for example, a company must sign a formal agreement to share any of its IP that might be required to implement PCI products. Such agreements represent substantial concessions by the original innovators, who give up their right to future royalties in order to create a larger market for the standard. These agreements are important to later adopters, who can develop standards-based products without fear of being sued later by fellow members.

No such explicit contract, however, exists among JEDEC members. Industry sources tell me there have been many cases of major companies participating in JEDEC standards efforts only until critical patents issue—and later refusing to sign licenses for the necessary technology, or demanding licensing fees out of proportion to the technology's actual value. No matter what happens with this case, JEDEC should consider implementing more-binding agreements among its members.

Rambus left JEDEC in June 1996, saying "Rambus plans to continue to license its proprietary technology on terms that are consistent with the business plans of Rambus, and those terms may not be consistent with the terms set by standards bodies, including JEDEC." The company's decision to leave JEDEC may have been influenced by a May 1996 consent decree against Dell by the U.S. Federal Trade Commission, in which Dell was prohibited from enforcing a patent it received after it failed to disclose the pending application to VESA, another standards organization.

By 1996, however, JEDEC's SDRAM standards were essentially complete. It took another three and a half years for the related Rambus patent applications to wend their way through the Patent Office, but once those patents issued, Rambus began approaching SDRAM vendors about licensing them.

Regardless of the moral quality of Rambus's behavior in JEDEC, the courts may find that the company's behavior was legally acceptable. Such a ruling would put Hitachi in a difficult position, because its other defensive arguments, especially those against the validity of the patents, do not seem to be especially strong.

Hitachi argues, for example, that Rambus is in violation of the Sherman Anti-Trust Act, because it is attempting to establish a monopoly over the intellectual property required to make SDRAMs, which constitute the majority of DRAMs on the market today. Monopolies established by superior technical innovation alone, however, are not illegal.

The patents themselves appear substantial enough. Rambus makes no claim to the notion of synchronous memory interfaces per se, only to significant enhancements required for practical products. In addition to the four U.S. patents listed in the original complaint, Rambus has since added four more: patent numbers 6,032,214; 6,032,215; 6,034,918; and 6,038,195. (Only the latter two are mentioned in Rambus's complaint to the ITC.)

The courts may never rule on the merits of these or any other arguments, however. Cases such as this one often conclude in a negotiated settlement. I suspect Rambus and Hitachi will be willing to settle their differences if Rambus withdrew its claims over SDRAM and Hitachi agreed to produce and promote RDRAM.

I believe Rambus had an obligation to disclose its relevant patent applications during development of the SDRAM standard. Because it withheld this information, I believe Rambus gave up its right to assert those patents with respect to SDRAM. I also believe this principle extends to the use of the same features in later SDRAM implementations such as PC133 and DDR SDRAM.

Some of Rambus's patents were never relevant to the original development of SDRAM technology, however, notably those related to double-data-rate signaling. As long as these patents can stand on their own merits, Rambus is entitled to license them as it sees fit for more-modern memory devices, including DDR SDRAM and anything that comes of the ADT effort.

Like any company participating in a cooperative market, Rambus has certain obligations—but it also has rights. The patent system, in particular, exists to protect intellectual-property rights. Any company that expects legal protection for its intellectual property must accept that Rambus is entitled to the same protection. Other considerations, such as the size and cost-sensitivity of the memory market, simply are not relevant.

We created and empowered our government for one reason alone—to protect our rights. Ultimately, all rights derive from one right—the right to own and control property. We must not sacrifice this right in pursuit of some temporary and debatable improvement in the pricing or availability of some mere commodity.