Litigation-happy Rambus, which specializes in high-performance memory interfaces, has taken on technology giant IBM again, despite a court ruling that said IBM had not infringed on the company’s patents. Rambus, which derives the majority of its annual revenue by licensing patents, claimed the U.S. patent office ruling was in error, according to a Reuters report. Rambus obtained the patent in 2002. “The board committed errors of fact and law in its orders, decisions and judgment,” Rambus said in its complaint, according to Reuters.
According to the report, the patent in question concerns “a memory controller that communicates to at least one memory subsystem, with an independent point-to-point link used between the controller and each subsystem,” outlined in the U.S. Patent and Trademark Office’s Board of Patent Appeals and Interferences ruling on June 24.
Earlier this month, Rambus announced it signed a patent license agreement with chip maker Nvidia related to certain memory controllers. Under the agreement, Rambus granted Nvidia a patent license for certain memory controllers at a 1 percent royalty rate for SDR memory controllers and a 2 percent royalty rate for other memory controllers, including DDR, DDR2, DDR3, LPDDR, LPDDR2, GDDR2, GDDR3, GDDR4, and portions of GDDR5 memory controllers. Nvidia granted no licenses to Rambus.
The agreement followed a patent-infringement suit begun by Rambus against Nvidia in 2008. According to the U.S. International Trade Commission, which July 26 ruled that certain products containing the Nvidia chips may not be imported, Nvidia infringed on three patents held by Rambus. The ITC ruling pertained to Nvidia’s GeForce, Quadro, nForce, Tesla and Tegra chips, which are used by customers such as Hewlett-Packard, Asustek Computer and Biostar Microtech. After the ruling, Nvidia officials said the company would pay a bond that will enable it to continue supplying HP and other OEMs with chips while the company appeals the ruling.