Richard Gibbons of TMF wrote an article which mentioned Rambus in an unfavorable light. I wrote a post where I stated the longstanding desire of TMF Rambus Discussion Board for TMF to publish a story of substance about the Rambus saga. Richard suggested I write such a piece myself and submit it for publication to TMF. I did so, and heard from Richard today that TMF was not interested in having a guest writer. Therefore, I am posting the article here. Please understand that I was writing for an audience that might not know anything about Rambus. Also, I was attempting slightly to approximate the "lighthearted" style that seems to appeal to TMF (see last sentence). It had no title, but I'll title it here anyway. DO YOU GET MY THRUST?Can you imagine a company that could increase its forward revenues 10 times over in a matter of months? What stock out there also has the potential to recoup 6 or 7 years of “lost” revenues in one court decision? Do you know of a company that can drop 80% of its revenues to the bottom line? What small company do you know of that has invented, patented, and licensed the technology that uncorked the “bottleneck” between CPU and Memory, but has had most of that technology co-opted by huge memory maker companies who joined together and went to war rather than pay licensing fees for the technology? The answer to all of the above is Rambus, Inc. Every now and then a stock story rolls around that is so unique, compelling, and so well-positioned for rich rewards that it demands an investigation. Such is the complicated, Byzantine, mysterious, and heart racing tale of Rambus Inc. The (ram) bus-ride began in 1988 with a conversation over dinner between two brilliant EE professors, Mike Farmwald and Mark Horowitz. Their vision, mapped out on a napkin that night, would revolutionize computer communications. CPUs were rambling along at 20 million cycles per second at the time. Professor Farmwald suggested that he could make them run at 500 million cycles per second. While CPU speeds were increasing rapidly, main memory (DRAM) speed was unable to keep pace. Over the next couple of years Farmwald and Horowitz solved the “performance gap” with a combination of inventions featuring “delay time”, “variable block size”, and “dual-edge clocking”. Thus began Rambus who in the intervening years has given us RDRAM in the Pentium 4 & Sony Play station 2, and more recently XDR DRAM and I/O connectors providing all the memory synapses in the “cell” by IBM, Sony, and Toshiba. In fact, Rambus's chip-to-chip connections are incorporated into PC main memory, hard drives, networking, consumer electronics (Xbox, HDTV, etc.), supercomputers, and even cell phones and automobiles. And for most of this they receive no revenue……........…yet. Why is this company of approximately 400 employees (mostly engineers) merely puttering along at roughly $20 a share? The answer is complex. It goes something like this. Most of the memory makers (eg.Samsung, Hynix, Infinion, and and Micron) loved the technology, but didn't want to pay for it. Rambus had revealed the secrets to them in ndas (non-disclosure agreements) around the time they filed their initial patent. What happened thereafter has been the subject of several lawsuits over the last six years. Finally last year, in #2 of a three phase trial, Hynix was convicted by a jury in San Jose of infringement of Rambus' patents on 37 counts. Earlier in the year Rambus was victorious in phase #1 trial on Unclean Hands. Just as phase #3 was to begin in August a salvo was launched at Rambus by the FTC causing postponement of the conduct trial. The FTC, at the behest of Micron and Hynix, had in 2001 brought antitrust charges against Rambus for allegedly subverted a standards setting organization (JEDEC) by supposedly tricking them into incorporating Rambus patents into their standards for SDRAM and DDR. The ensuing FTC trial before Administrative Law Judge (ALJ) Steven McGuire was the most extensive and expensive in the history of the FTC. The initial decision of 334 pages exonerated Rambus on every count, and even contained information which confirmed much of a parallel investigation going on in a sister agency, the Department of Justice (DOJ). The DOJ investigation of the DRAM makers found significant evidence of collusion and price-fixing by Samsung, Infinion, Hynix, Micron, and others in the industry. To date in this ongoing DOJ probe there have been record fines paid by these companies (over 800 million dollars). They confessed to conspiracy in the DRAM market (in some cases actually including RDRAM, aka Rambus Dram). Isn't it strange how one branch of the federal government, the DOJ, can be convicting of felony the same companies who meanwhile petitioned another branch of government, the FTC, to punish one of the victims of their conspiracy? Back to the August FTC salvo. It seems that after more than two years the FTC decided that their ALJ got it wrong in his ID. On appeal, they overturned the ruling that had cleared Rambus. They felt Rambus' failing to tell JEDEC of its impending patents didn't allow alternative work arounds for their technologies. Thus JEDEC incorporated them into their standards. The FTC finally issued their “remedy” for the supposed transgressions last week. There is little or no precedent for the FTC setting rates that can be charged by a private company, but they did so anyway. They told Rambus that they could only henceforth charge .25% for SDRAM and .50% for DDR (these technologies are being phased out), and 1% for connectors. Also, they reluctantly stated they had no mandate to set rates for DDR's subsequent iterations DDR2 and 3. Immediately Rambus stock price zoomed ahead 25% because not only was DDR2 exempted (next year the 30 billion dollar DRAM market is projected to be 80% DDR2), but also, the FTC tacitly conceded that Rambus' patents were valid by assigning a particular value to them. What does all this mean? On Friday of this week, February 16th, Judge Ronald Whyte will make some crucial decisions in his court in San Jose, California. If he denies Hynix the right to import the findings or results of the FTC into the Phase #3 trial and sets a date for such, Rambus may finally obtain resolution on past infringement. Having lost on Phases #1 and #2 Hynix has had to escrow well over 100 million dollars. If they lose phase #3 their goose is cooked. If Hynix loses, the remaining cartel members are likely to go down too, except Infinion which has settled with Rambus. Making matters worse for the memory makers is the possibility Friday that Judge Whyte sets the date for another trial, known as the DDR2 trial. This is a much stronger case for Rambus due to no FTC interference, and the fact that JEDEC & the memory makers could have worked around Rambus IP, but didn't. Rambus invented the technology for DDR2 long after leaving JEDEC. It's becoming increasingly probable that the cartel members must eventually seek settlement with Rambus. Currin Research figures that settlements would render Rambus in a position to reap approximately 1 billion dollars per year in revenues from a 30 billion dollar DRAM market. Further, Rambus connectors, while bringing lower rates, apply to a much larger market and can be expected to generate a similar return. Remember, the FTC said Rambus can charge 1% for their connectors. Rambus has the potential in 2007 to bring closure to their infringement litigation. If so, the (ram) bus-ride will slip into thruster gear such as it did in its early days when it topped $400 per share. As if all of this were not enough to confuse you there remains one monumental factor which I have not mentioned, the antitrust case. On April 12th there will come a Management Conference to set a date for the Rambus Antitrust Case against the memory maker cartel for their conspiracy against RDRAM. By the first week of March Micron will be required to release damning documents related to the DOJ investigation which will strengthen Rambus' already overwhelming antitrust case against those who allegedly colluded to destroy the company. This trial date will be zero hour for many executives, former executives, and perhaps accomplices in the press who in the early years were so brazen as to send emails and hold meetings flagrantly conspiring to undermine Rambus and drive them out of the market. One of the classics is from Linda Turner, Micron vice-president, who in June of 2001 states in an email to other Micron employees, we “....have actually been requesting Infineon, Samsung, and Hynix lower their DDR pricing to help it become a standard (and drive Rambus away completely).” If the resumption of the Phase 3 trial or the FTC remedy doesn't force the cartel members to settle with Rambus, maybe the embarassment and expense of being interrogated by Rambus attorney Joseph Cotchett in the antitrust case in Northern California will. That trial is on track to be heard in the summer of this year. This could be the year that the ram-bus finds its thrusters again. All aboard!
Best Of |