“The fact is that one side thinks that the profits to be won outweigh the risks to be incurred, and the other side would rather avoid danger than accept an immediate loss.” 
 

--Thucydides
 
The Greek historian and author of The History of the Peloponnesian War, Thucydides could have been peering at the future Rambus versus DRAM cartel battle. 
 
The cartel knows it is guilty.  However it has been playing out a series of potential legal escape hatches on its way to the inevitable.  They have avoided the “immediate loss” of a royalty agreement settlement.  But we believe once the loss becomes more imminent they would rather avoid billions of liability danger and will indeed reach settlement.
 

The cartel members (except for Micron who received amnesty from the Department of Justice as a cooperator) have paid fines and had employees serve some jail time for price-fixing in the DRAM market.  Rambus contends that RDRAM was also a target of the price fixing and that, but for the price fixing, Rambus would have enjoyed substantial royalties on RDRAM. 
 
There is a mountain of evidence to support the Rambus case.  It’s the kind of evidence that would lead a sane defendant, operating in a slim to no margin business, to avoid the catastrophic result of losing this case to Rambus and seek a settlement.  Or as Thucydides may have posited substituting Rambus for Athens… “pay a royalty tribute to Rambus and thus survive, or fight Rambus and be destroyed”. 
 
Certainly neither Hynix nor Micron could incur the damages Rambus seeks and remain viable.  The DRAM cartel members face the possibility of losing not only the huge antitrust case but also being on the hook for future royalties because of infringement.  Rambus likely would settle the litigation if the royalty issues were resolved.
 
DRAM cartel members cannot afford to lose the case
 
Going to trial when you know you are guilty is not such a good idea.  This is especially so with treble damages facing your already hamstrung balance sheet.  Throw in smoking gun emails and worried statements concerning coordinated plans to attack RDRAM, that incredibly include the phrase “please visit me if I end up in jail”,  and one is faced with reality that avoiding the danger of an immediate loss is very likely if not inevitable. 
 
As hyperbolic as the please visit me in jail line may appear, that’s exactly the place from which several price fixers would later give their depositions to Rambus….from behind bars.  All in all it’s not looking good for the price fixing team aiming to defeat RDRAM.  
 
Just How Guilty are They?
 

A look at the facts appearing in court documents tells the tale.  Besides making one ponder the rhetorical, “who is stupid enough to document criminal acts in email?”, one is left with the question of why has this case not already resulted in a settlement.  However with the strong evidence against the DRAM manufacturers in plain sight, Rambus investors should feel a sense of confidence about the outcome if it actually goes to trial.
 
Rambus’s case essentially boils down to a question of whether price fixing and other anticompetitive collusive acts in the DRAM industry damaged RDRAM’s success in the marketplace.  The damage to Rambus is royalties it would have received on RDRAM sales. 
 
The answer to the question is a simple and clear yes. 
 
Below is a review of some of the more damaging information one can hope to find before heading into court.  Indented passages come directly from court documents.
 

Did DRAM manufacturers fix prices?

Yes. 
 
Back in the summer of 2002, the United States Department of Justice initiated a criminal investigation with respect to price fixing in the DRAM industry.  Micron took part in the DOJ amnesty program.  Micron has publically admitted that Micron employees engaged in the price fixing.  Infineon pled guilty to participating in a criminal conspiracy to fix prices and paid a fine of $160 million.  Hynix pled guilty and paid a fine of $185 million.   
 
In the Hynix plea agreement the company agreed to cooperate with the U.S. government in the federal investigation.  Several months later Samsung pled guilty to the price fixing conspiracy, including price fixing on Rambus RDRAM, and paid a fine of $300 million.
 

So the main defendants in the case, Samsung, Hynix and Micron are all admitted price fixers.  
 
 

Did Price Fixing Harm RDRAM?
 
Yes.
 
Samsung’s admission that it fixed prices on RDRAM is a critical admission because it is well established that the reason RDRAM failed to gain more market penetration, and ultimately was a niche product primarily in SONY Playstation systems, was because the price of RDRAM was high compared to alternatives. 
 

Targeting RDRAM
 
More damaging than the simple act of fixing prices is the evidence that the manufacturers coordinated prices specifically in order to thwart RDRAM. 
 
RDRAM was a threat and an enemy from the beginning.  Any manufacturer acting alone could not thwart it though that at least would have been legal.  A collusive conspiracy against a competitor is a crime…and that was the path chosen by the defendants.
 

The RDRAM threat
 
Intel had announced that it intended to use Rambus memory in its roadmap.  This caused much concern among DRAM manufacturers especially Hynix.
 

In September 1996, Farhad Tabrizi, Hynix's Worldwide Vice President of
 

Marketing, predicted: 
 

If Intel implements Rambus all other applications will move that
 

Direction to leverage the strength and volume of the PC market. . .
 

I urge you to please educate others and get their agreement to say
 

"NO TO RAMBUS AND NO TO INTEL DOMINATION."           
 

            ----------- 
 

At a January 1997 meeting of the Synchlink Consortium (also attended by
 

DRAM manufacturers Hynix, Micron, Samsung, and Infineon, among others),     a Hynix representative stated: 
 

[The Rambus-Intel relationship is] a doomsday scenario for
 

DRAM business: If Intel/Rambus [is allowed to] control all IP,
 

DRAM suppliers will be nothing more than foundries, with profits
 

going into Rintel's[ Rambus+ Intel's] pockets.
 
Thus, long before Rambus had a single DRAM device in an Intel-based PC, the idea of thwarting Rambus was hatched as a cooperative plan. 
 
In hindsight the DRAM makers have done little to avoid their concerns.  Today DRAM market margins are besieged by commodity pricing and oversupply.  Some manufacturers are not quite foundries but perhaps even worse off than one that makes non commodity products.  In addition, their commodity alternatives to RDRAM infringe Rambus patents. The path they chose left them pleading guilty in the DOJ case, facing a Rambus antitrust suit, and facing Rambus infringement damages to boot.
 

The Plan Required Rambus RDRAM to fail with Intel
 
Because Intel chipsets dominate the PC market, whatever DRAM is compatible with those Intel chipsets can become a de facto standard.  Intel chipset volume simply drives the lowest cost of the associated DRAM. 
 
With Intel’s endorsement of Rambus, RDRAM was on its way to that status.  But little did Rambus and Intel know that the industry, desperately seeking control, would resort to collusion to thwart Rambus and Intel.   
 

15. Between 1997 and 1999, the Defendants undertook concerted efforts to lay the groundwork necessary to keep the price of Direct RDRAM high and supply
 

low once Intel's RDRAM-enabled chipsets launched in late 1999, knowing that,as Desi Rhoden (of IBM) put it, "Intel won't change course unless Rambus fails."
 
Behind the scenes the manufacturers were collectively working “to say no to Rambus and Intel domination.”
 

Setting RDRAM Prices High

It’s hard to have a better sponsor in the PC market than Intel.  The DRAM makers had their work cut out for them.  RDRAM had to fail.  The manufacturers demonstrated their zeal to accomplish the task.

In efforts to damage RDRAM in the marketplace of the press, a coordinated effort to smear RDRAM was undertaken. 
 

A Micron marketing executive named Jeffrey Mailloux informed Hynix Vice
 

President Tabrizi on February 20, 1998 that he had called a reporter and had told him that RDRAM was "at least 30 % more expensive to manufacture than SDRAM.  Mr. Mailloux urged Mr. Tabrizi to make a similar call to the reporter, warned Mr. Tabrizi not to forward his email to anyone and asked Mr. Tabrizi to "please visit me if I end up in jail. . . ."
 

Apparently Mr. Mailloux was aware that such coordinated actseamong competitors seeking to cooperate to harm a third party could lead to criminal punishment. 
 

Hynix Vice President Farhad Tabrizi reported that Hynix presented Intel with
 

Inflated projected prices and production volumes for RDRAM and "encouaged
 

every DRAM manufacturer to do the same in order to let Intel not generate Rambus oversupply.
 

There are many more examples of the specific actions taken in response to a perceived threat of Intel/Rambus domination.  All of the actions essentially boil down trying to cripple RDRAM and its success within Intel.  Some examples include:
 
 

·         Keep the price of RDRAM high. 
 

·         Cause trouble with the Intel/Rambus, OEM relationship by coordinating false production values and high price expectations.
 

·         Falsify press stories in a coordinated manner  
 

·         Exaggerate RDRAM costs with coordination
 

·         Limit the production of RDRAM despite making inflated production promises.
 

·         Provide inflated cost estimates for RDRAM production in order to influence OEM roadmap decision making.
 
When you work together in such an enterprise of activity it’s called a conspiracy.  And that folks is exactly what Rambus can show.
 
Suppliers Needed to Work Together
 
The suppliers needed to work together.  Thre was often concern that a supplier had broken ranks and “joined the dark side” (Rambus).  The only way an output restriction works is if the parties don’t cheat and simply take more of the Total Available market (TAM) while another cartel member idles back production.
 
 

By May 1999, with the official launch of RDRAM coming closer, Hynix
 

Employees and sales executives were discussing a plan to keep Hynix's RDRAM prices high, but they recognized the need to "get the rest of the suppliers to do the same" in order for the scheme to be effective. 
 

----
 

Recommendation by Hynix manager Andy Ha that Hynix "get together with Samsung and ask them to suggest to Micron to have a joint management meeting for price control in the future.") 
 

---- 
 

On July 20, 1999, the Hynix Marketing Manager responsible for RDRAM,
 

Mario Martinez, proposed to a group of senior Hynix managers that Hynix meet with Samsung and "work with them to limit the supply [of RDRAM] in the market." 
 

Mr. Martinez explained:" With Samsung building significant amounts of product, we need to work with them to limit the supply in the market, otherwise we both will be competing for market share which will result in an oversupply. We have to meet with Samsung and discuss our and their production plan, TAM analysis, and targeted market share(WW and Strategic Account) and reach consensus." 

---- 
 

A Hynix manager based in Korea responded that he had a "connection in Samsung” and that Samsung "actually" had the "same idea for Rambus business compare with you." 

The above passages show that Hynix had every intention of limiting the production of RDRAM so that it would fail.  They planned to work with Samsung and did.  And Samsung has admitted to RDRAM price fixing.  In the investigative sense these communications are called smoking guns. 
 
Per se Illegal
 
Working with suppliers to limit supply is a per se illegal antitrust offense.  The limited supply creates the high price.  The high price (and limited supply) ultimately caused Intel to move away from RDRAM because of concerns from OEMs like Dell and Compaq.  The suppliers responsible for the illegal acts are responsible for the resulting harm.  The harm to Rambus was the vast reduction in RDRAM sales and royalties Rambus would have received.
 
Limited Early Success
 
The manufacturers 1998-9 plans were only partially successful.  RDRAM was saddled with a high price perception in the industry, Intel continued its support.  OEMs were also planning RDRAM based systems. 
 
So price fixing to thwart RDRAM continued into 2000.  Samsung became the company that set the price that was to be the low limit.  Manufactures were to check with Samsung and not quote RDRAM below these levels. 
 

59. In a February 16, 2000 email, another Micron sales representative reported that he had just talked to someone from Samsung who said that Compaq is pressing hard for Rambus support "and that Compaq had made a "sooner than  expected technology switch (towards Rambus) based on the apparent success Dell is having."

60. In response to customer demand for RDRAM, the Defendants chose collusion rather than competition.  A February 8, 2000 email by Hynix sales representative (Mr.) Byrd, for example, reported that he had learned from Samsung that Infineon "apparently did a stupid thing on RDRAM pricing" to Compaq by offering a "lower price than Samsung” 
 

---- 
 

Infineon's "stupid" mistake was in "quoting pricing without first checking with
 

Samsung” 
 

---- 
 

The Samsung representative also told Mr. Byrd that unlike Infineon, NEC had "checked with" Samsung and was "in line with [Samsung] pricing;' 
 

---- 
 

Hynix and Samsung hold regular meetings in spring 2000 to discuss the respective companies’ “Rambus status” and “production plan," as memorialized in a March 30, 2000 Hynix memorandum. 
 

---- 
 

The Hynix memorandum listed two Samsung employees who were acting as
 

"contact person[s]" with respect to RDRAM issues who would be "contacted
 

Whenever there are some issues by voice or meeting….The memorandum also 
 

stated that "if possible [Hynix and Samsung] will have a meeting month after month."
 

The close relationship between Samsung and Hynix in keeping RDRAM prices high is telling.  Samsung stood to gain if RDRAM prices remained high.  That is, even if RDRAM was limited to SONY Playstations, as the dominant supplier of RDRAM, Samsung would make nice profit margin on the niche product.  Samsung would become the major RDRAM producer as well as a major DDR producer with an RDRAM margin cushion. 
 
Hynix simply wanted the industry wanted out of RDRAM altogether.  (Remember say no to Rambus and Intel domination).
 
Since RDRAM had certain design wins, Samsung would benefit from RDRAM no matter how much it sold.  Hynix had the clear motivation of making sure Rambus failed.  The threat on Intel/Rambus domination was their motivator. 
 
Examples of RDRAM collusion and price fixing coordination among the defendants are too plentiful to mention.   The DOJ documents cite hundreds of instances of interaction.
 
Coordinated efforts to discourage Dell and Compaq from using RDRAM also occurred.  Hynix refused a request from DELL for more RDRAM.  Hynix even refused to respond to a request for output increase. 
 
Because the OEMs were getting a fairly consistent coordinated message from suppliers that RDRAM prices would remain high, they became concerned about RDRAM systems for the future roadmap.
 

But RDRAM is needed unless you have something “acceptable”…it’s not SDRAM and DDR is not quite ready
 

The trouble for the manufacturers bent on stopping Intel/Rambus domination was that Intel was moving ahead with Rambus for technical reasons.  In early 2000 Intel launched the Pentium 4 which was designed specifically to take advantage of RDRAM bandwidth. 
 
Despite some frustrated rumblings of the OEMs at the price points for Rambus based Pentiums by this time, Rambus was still well established on the Intel roadmap. 
 

However by early 2001 an ominous sign appeared in a slide at the Intel Developers Forum concerning the roadmap.    DDR appeared as a “possible solution” with the caveat
“Competitive RDRAM pricing minimizes the roll of  DDR”. 
 
This is an important caveat  as it clarifies the entire Rambus claim for damages.  Intel’s own slide reveals that DDR was aided in usurping RDRAM in the eventual roadmap by non-competitive prices on RDRAM.  With competitive pricing, RDRAM would have been the marketplace winner and de facto memory standard for the industry. 
 
Because the DRAM market participants had obviously fixed prices on RDRAM high, DDR was eventually allowed to get the leg up.  Although Intel also developed a Pentium 4 using SDRAM, SDRAM was never really capable of meeting the Pentium 4 performance needs. 
 
DDR was still emerging albeit with lots of problems.  The fact that the DDR design borrowed substantially from Rambus patents and methods on display in RDRAM was of little concern to the DRAM makers.  Their goal of “killing RDRAM” was spelled out early on and implemented quite well. All that was really required for DDR to eventually replace RDRAM in Pentium 4 was a continuation of non-competitive RDRAM pricing.
 
Continued cooperation in price fixing
 
Still, by early 2001 DDR was not yet ready to meet the Pentium 4 demand.  The manufacturers needed more time for DDR and continued in efforts to thwart RDRAM after the launch of the “RDRAM only” Pentium 4. 
 
 

 This joint strategy was explained in a June 5, 2001 email written by Micron
 

Managing Director of International Sales Linda Turner.  Ms. Turner was responding to reports by her staff that Hynix was lowering its DDR pricing.  
 

Rather than view this price competition with alarm, Ms. Turner explained how low DDR price projections could tip the market away from RDRAM: 
 

No problem! We want DDR to explode into the marketplace so
 

have actually been requesting Infineon, Samsung, and Hynix to
 

lower their DDR pricing to help it become a standard (and drive Rambus away completely).
 
 
This is another smoking gun email.  “we have been requesting Infineon, Samsung and Hynix to lower their DDR pricing to help it become a standard (and drive away Rambus completely).  This is essentially what Rambus charges in their complaint...the price fixing cause and effect relationship for the success of DDR over RDRAM.
 
While a natural competition between DDR and RDRAM would not have been a problem, the collusive price fixing is illegal and the actual cause of DDR success.  It’s even a stated goal memorialized in the defendants email. 
 
The manufacturers were able to continue to attack RDRAM with price fixing and output restriction as DDR received a coordinated price cut. 
 
Too Greedy Too Fast
 

The manufacturers were quick to raise DDR price once they thought they had RDRAM “killed” but that premature increase in DDR pricing led to resurgence in RDRAM. This “problem” (a resurgence of RDRAM) was attacked head on with more price fixing by keeping a lid on DDR prices until the job of thwarting RDRAM in the PC market was complete.  
 

The Defendants nevertheless  became concerned in January and February 2002 that their coordinated price increases on DDR SDRAM would lead a renewed marketplace interest in the RDRAM device.  For example, an email by a Micron sales manager warned that because of  "price premium of DDR over SDR(AM), "orders for chipsets using RDRAM had increased at both Dell and Intel. 
 

A January 16, 2002 news article entitled "DDR Fumbles RDRAM Scores" noted that "rising DDR prices spur Rambus resurgence." …The article pointed out that "swiftly rising"  DDR prices had called  "into question the validity of tech articles written when you could buy 512M B of DDR for under $60. . . ."  
 

The article concluded: 
 

In the rush to the DDR Promised Land, we may have lost track of the reason we
 

started.  Price/performance as of today, RDRAM gets a higher rating in this regard. . .
           
            ----
           
 

In light of this "threat" caused b y "the DDR/SDR price gap, "Micron's
 

Worldwide Vice President of Sales Mike Sadler asked a direct report, Bill Lauer, to "educate" other DRAM manufacturers about the "virtues" of price parity between SDRAM and DDR.
 

So what the cartel did next is documented as lowering the price of DDR closer to that of SDRAM through collusion with the specific target of eliminating RDRAM as a resurgent threat.  This basically worked though it resulted in lots of red ink for the manufacturers.
 
Part of the price action started with response to Rambus optimism that DDR pricing was opening a window for RDRAM.  In the January 2002, “scary message” memo Micron employees discuss the pricing parity issue as related to a Rambus threat.
 

By March 2002 the pricing parity and price fixing in response to RDRAM resurgence was enacted.
 

  
While the final blow to RDRAM came at great expense to the manufacturers as they were selling below cost for DDR, it got the job of killing RDRAM done.  Ultimately Intel moved away from the RDRAM chipset and DDR and later DDR2 became the mainstream PC devices.
 
The Fat Lady is Warming Up
 
The DRAM cartel has a big problem with this case.  They are dead on arrival versus the complaint. 
 
Rambus investors should take heart that the day of reckoning for the criminal acts of the cartel is near.  The court case is scheduled to begin in September of 2009.  While Rambus has had a long road of proving patents are infringed due to standard setting issues and document spoliation among other things, the deeds of the DRAM cartel are laid bare enough even in this not so cursory overview to see that RDRAM was the victim of a criminal conspiracy.  
 
Samsung’s Shame will be Rambus’s gain
 
Samsung has admitted to fixing prices on RDRAM.  Multiple executives
 
 

Il Ung Kim, Vice President of Marrketing
 

Young Woo Lee, Sales Director
 

Sun Woo (“Sunny”) Lee, Senior Manager of DRAM Sales
 

Yeongho Kang, Associate Director, DRAM Marketing
 

Young Hwan Park, Vice President of Sales
 

Thomas Quinn, Vice President of Marketing for Memory Products
 
pled guilty to a criminal conspiracy with respect to DRAM price fixing.
 
We believe Samsung has interest in settling litigation with Rambus.  It has the much to lose in the litigation.  Samsung’s competitors are already reeling from the rigors of true competition with Samsung in the post price fixing arena.  Samsung should be concerned that Micron would leave them holding the liability bag just as occurred with the DOJ. 
 
Similarly, Micron and Hynix have perhaps the most to lose as they certainly would be ruined by a huge antitrust damages award.  
 
All of these elements point to the reality that litigating in the face of a mountain of incriminating evidence is simply not in the best interest of the defendants.
 
Justice
 
Rambus is providing a path to peace.  Settlement and licensing.  Rambus’s proposed licensing scheme in the EU antitrust settlement points to a reasoned approach to a path forward.  The DRAM makers get by with a royalty rate essentially in line with what it would have been with RDRAM. 
 
While Justice delayed is justice denied, sometimes its justice nonetheless.
 
 

On Dilution
 

The recent debt offering by Rambus allows Rambus to repurchase the debt anytime after 2012.  Given that Rambus should have either patent damages or perhaps settlement by that time we view the debt offering as not too worrisome.  If Rambus is more than 30% above the $14 and change value it will likely be very much more than that by then and the debt offering will have been worth the time insurance. 
 

The market has discounted dilution while ignoring the binary effect of an end to the litigation.  In other words it is difficult to price the dilution impact but with settlement investors will not be displeased with a bit of dilution. 
 

Our basic view is that the debt offering gives Rambus the ability to ride out this litigation to its amenable or non amenable end game.  Because the liability is so large compared to the debt incurred we view it as tangential to the actual value of shares.  The real wildcard is the antitrust case and its potential to induce settlement.   
 

Though once a defendant in antitrust itself, Rambus will almost certainly have the last laugh in antitrust.  
 

Rambus is a Fredhager.com portfolio stock
 

Rambus is a CurrinResearch.com portfolio stock
 

Thank you.