Traditionally, District Court judges have issued permanent injunctions to stop infringement of valid patents, regardless of any pending reexamination of the patent in question. Only if the rejection of the patent claims in reexamination is “final,” would the courts refuse to enjoin the accused infringer. A few judges – notably Judge James Cohn of the Southern District of Florida and Judge Sue Robinson of Delaware – have recently shown a willingness to defer a decision on injunctive relief even where the rejection of the patent in reexamination is not final.
The Rambus/NVIDA investigation raises the question of the extent to which the ITC is willing to order relief on the basis of patents that are in reexamination at the PTO. Unfortunately for NVIDIA, its argument to the ITC that it stay relief against NVIDIA was weak because its attack the Rambus patents by reexamination has been “an incomplete success” (to borrow a phrase from President Carter).
That weakness may account, in part, for NVIDIA’s agreeing at the end of last week to a license to the Rambus patents, effectively concluding the dispute between the parties. Still, the ITC’s analysis in the Rambus/NVIDA investigation remains of great interest to the IP community, as it suggests that the ITC will be extremely reluctant to withhold relief against infringing imports on the basis of an uncompleted reexamination.
In his Initial Determination, the ITC’s Administrative Law Judge Theodore Essex found that three Rambus patents – the so-called “Barth I patents” – were valid and infringed by imported semiconductor chips having synchronous DRAM controllers from NVIDIA. On the other hand, he found two other Rambus patents – the “Ware patents” – to be either anticipated or obvious.
This past week, the ITC affirmed Judge Essex determination in each respect, despite questions raised regarding the validity of the Barth I patents in pending reexaminations at the PTO. Having found that NVIDIA’s products infringed valid patents, the ITC issued a limited exclusion order prohibiting entry into the United States of NVIDIA’s products covered by the Barth I patents, and an order requiring NVIDIA (as well as other Respondents) to cease and desist from the sale of infringing NVIDIA products already in the United States.
The current status of the five Rambus patents, both at the ITC and at the PTO, is summarized in the table below.
|Rambus Patent No.||Rambus Patent Title||Determination of the U.S. ITC||Status of Claims in Reexamination|
|6,591,353||“Protocol for Communication with Dynamic Memory”||Claims 11-13 are not anticipated or made obvious by Farmwald ‘755.||Claims 1, 5, 7, 11, 14, 19 & 23 anticipated by Hayes; patentability of Claims 2-4, 6, 8-10, 12-13, 15-18, 20-22 & 24-26 confirmed – on appeal to the PTO Board. NVIDIA has asked that prosecution be reopened because of inconsistent claim construction by the examiner.|
|6,470,405||“Protocol for Communication with Dynamic Memory”||Claims 11-13, 15 & 18 are not anticipated or made obvious by Farmwald ‘755.||Patentability of Claims 1-37 confirmed – on appeal to the PTO Board.|
|7,287,109||“Method of Controlling a Memory Device having a Memory Core”||Claims 1-2, 4-5, 12-13, 20-21 & 24 are not anticipated or made obvious by Farmwald ‘755.||Claims 1-25 rejected by Examiner as anticipated by Farmwald ‘037 – on appeal to Board where NVIDIA argues the claims are invalid on the basis of other prior art, as well.|
|7, 210,016||“Method, System and Memory Controller Utilizing Adjustable Write Data Delay Settings”||Claims 7, 13 & 21-22 are anticipated and made obvious by Coteus.||Claims 1-4 & 7-24 rejected by Examiner as anticipated or made obvious by Coteus and Sato – both Rambus and NVIDIA have commented on the rejection and are waiting for the Examiner’s response; patentability of Claims 5-6 is confirmed.|
|7, 177,998||“Method, System and Memory Controller Utilizing Adjustable Read Data Delay Settings”||Claims 7, 13 & 21-22 are anticipated and made obvious by Coteus.||Claims 1-4 & 6-24 rejected by Examiner as anticipated or made obvious by Coteus and Sato – both Rambus and NVIDIA have commented on the rejection and are waiting for the Examiner’s response; patentability of Claim 5 is confirmed.|
Significantly, NVIDIA argued that the ITC should stay any remedial order because “with almost complete certainty,” the reexaminations at the PTO would result in cancelation or amendment of the Barth I and Ware patent claims. The ITC’s issuing remedial orders now would be a waste of judicial resources since the Commission will need to change those orders at a later date, NVIDIA argued.
But the ITC was not buying. First, the reexaminations “are not yet final.” Second, the PTO has, so far at least, found some of the asserted Barth I claims to be valid over the prior art, in agreement with the ITC. And even if the ITC did agree with NVIDIA on the suitability of a stay, withholding “the exclusion order with respect to those claims that the PTO has preliminarily determined are invalid would still result in exclusion of an identical set of accused products, because all accused products were found to infringe all the asserted Barth I claims.”
One wonders whether NVIDIA’s assertion – that the ITC should refrain from issuing an exclusion order or a cease and desist order pending completion of the reexaminations – might have been stronger if it had been based on the “public interest,” specifically, that the public interest would be harmed by issuance of such orders based on patents ultimately held to be invalid.
It is possible that the ITC will always exclude imports that infringe a valid patent, unless the rejection of that patent by the PTO is final, regardless of any future change in District Court practice. The ITC has authority to issue only prospective relief. It does not have the option, enjoyed by the courts, of compensating the patentee by awarding it money damages in the event that the patent is ultimately determined to be valid. Moreover, the ITC must consider the opinion of Congress that might regard a decision to withhold relief, in cases such as this, as a failure to enforce the country’s trade laws.
Still, it seems likely that NVIDIA’s argument was doomed in any event by PTO’s finding, at this point in time, that at least some of the claims are valid over the prior art. Although NVIDIA has succeeded against the Ware patents, both at the ITC and the PTO, it has so far failed to establish in either forum the invalidity of certain claims of the Barth I patents.