By Dennis Crouch
Marvell's stock price plunged about 10% following news that a jury had awarded well over $1 billion patent infringement verdict to Carnegie Mellon University. Download MarvellVerdict. The verdict is based upon finding of direct literal infringement of claim 4 of Patent No. 6,201,839 and claim 2 of Patent No. 6,438,180. The CMU hometown Pittsburgh jury also found that Marvell induced infringement by encouraging customers to use its chips. Regarding validity, the jury held that the asserted claims were neither anticipated nor obvious. One of the key pieces of evidence in the trial was an old email from a Seagate engineer noting that the CMU invention went well beyond the current state of the art. (Of course, that email was focused on advances found in the disclosure rather than the broad claims asserted by CMU.) Seagate was apparently happy to provide that document to be used against its market competitor. It is that same Seagate engineer's prior patent application that served as the core of Marvell unsuccessful invalidity argument.
An interesting element of the verdict involved particular willfulness questions. The jury particularly held that (1) Marvell knew of the asserted patent before the lawsuit; (2) had no reasonable defense for its actions; and (3) knew or should have known that its actions were infringing. The verdict did not, however, ask whether the action was "willful." CMU will ask the court to rely upon those answers to treble the damage award. If the damage award is trebled (tripled) and interest added, the award will be roughly equivalent to Marvell's $3.8b market valuation. (The award is larger than CMU's current endowment.) In the initial complaint, CMU also requested attorney fees and a permanent injunction.
This is the second billion dollar judgment against a Quinn Emanuel client awarded in the past few months.
Marvell has announced that it will prevail on appeal. In a pretrial ruling, district court judge Nora Fischer denied Marvell's motion for summary judgment of anticipation, but noted that it was a "close call" based upon her construction of the asserted claims. Marvell did not exercise its authority to seek a reexamination of the patent at the USPTO. Over the next six months, both parties will file and respond to post-verdict motions. Judge Fischer announced her intention to make a final ruling in the case in May 2013. If the verdict is confirmed by Judge Fischer, Marvell will then have an opportunity to appeal after posting a "supersedeas bond." If the damage award is roughly equivalent
The asserted patent claims are basically signal processing logic algorithms for determining the value of items coming from a computer memory signal. This is necessary because the "0" and "1" that we normally talk about for binary digital signals is not actually accurate. In particular, the invention indicates that you should apply a "signal dependent" function to calculate the value as a way to overcome noise in the signal. Apparently a key distinguishing factor from the prior art is that the decoding functions used are "selected from a set of signal-dependent functions." The asserted claims do not identify the particular functions used, only that functions are used.
The two infringing claims are as follows:
Claim 4 of US Patent 6,201,839:
A method of determining branch metric values for branches of a trellis for a Viterbi-like detector, comprising:
selecting a branch metric function for each of the branches at a certain time index from a set of signal-dependent branch metric functions; and applying each of said selected functions to a plurality of signal samples to determine the metric value corresponding to the branch for which the applied branch metric function was selected, wherein each sample corresponds to a different sampling time instant.
Claim 2 of US Patent 6,438,180 (Claim 2 is dependent from Claim 1 and I combined the two together):
A method of determining branch metric values in a detector, comprising:
receiving a plurality of time variant signal samples, the signal samples having one of signal-dependent noise, correlated noise, and both signal dependent and correlated noise associated therewith; selecting a branch metric function at a certain time index; and applying the selected function to the signal samples to determine the metric values, wherein the branch metric function is selected from a set of signal-dependent branch metric functions.
Folks focused on subject matter eligibility issues will notice the obvious problems raised by the asserted claims.