linux.com :: dram pricing: the fix is in

linux.com :: dram pricing: the fix is in linux.com search search log in  |  create account  |  submit story articles case studies features news newsvac reviews documentation what are linux howtos? where can i get linux howtos? howto translations categorized list of howtos single list of howtos single list of mini-howtos unmaintained howtos writing and submitting a howto copyright information distributions forums sponsor solutions open source and linux from hp perforce fast software manager about us what is linux? learn about linux download linux get linux help feeds features newsvac forums news video comments special offers get special offers on: linux application dev programming software email: feature dram pricing: the fix is in by on may 12, 2003 (8:00:00 am) print comments - by melanie hollands -price manipulation by manufacturers of pc memory components such as double data rate (ddr) dram is deliberate, periodic, predictable, and has been happening for decades. conveniently, dram price increases tend to be observed during critical periods, such as just before earnings reports or after periods of substantially subdued demand. it has also been suggested that some memory suppliers periodically withhold supply of dram and other memory components. this creates an artificial supply shortage, which in turn generates a level of artificial demand that drives up prices. irrespective of the cause of the price increase, dram prices are a key driver of pc-related and semiconductor stocks. as memory prices rise, so do semiconductor stock prices. the november, 2002 "rally" in semiconductor stocks was regarded by some to be the result of price fixing tactics. dram price increases propelled the pc-related, semiconductor and semiconductor capital equipment issues. ddr was at its highest level in over three months, which explains why micron technology stock rose nearly 40% to $17.30, up from october lows of around $12. and infineon traded up over the $10 mark, doubling from its october lows. while i believe that the demand for chips in the second half of 2002 was real -- it was part of the inventory build – it is the sell-through that remained a question mark. the build was likely stimulating demand at a time when most had written off a lift in real end-demand. the seasonal build in dram typically begins in the second calendar quarter and typically peaks mid-november. the 2002 price ramp didn’t last beyond that, and if history continues to be a guide i doubt that this year’s higher dram price trend will last beyond mid-november, but i am not making any guarantees. however, i do believe that it is largely these "supply shortages" that have supported these ramps in dram pricing and, in turn, semiconductor and related stock prices. heavily concentrated market dram supply is heavily concentrated in the hands of five major suppliers who, collectively, control nearly 82% of the worldwide market. in 2002, the top five suppliers were: samsung semiconductor, a division of korea’s samsung, the world's largest supplier of ddr, sdram (synchronous dram), and flash memory, and with >32% dram share, micron technologies’ subsidiary crucial technology (>18% share), infineon technology, subsidiary of germany’s infineon technology ag (12.8%) hynix semiconductor, from korea (12.8%), and nanya technology corp., of taiwan (5.5%). other dram suppliers include toshiba, mitsubishi, elpida memory inc. (a 50:50 japanese jv between nec and hitachi), and taiwanese manufacturer winbond electronics. more comprehensive dram market share details are provided in tables 1 and 2, below. in recent years, the computer memory industry has undergone significant consolidation, and continues to consolidate further. currently, half of total worldwide dram supply is controlled by two companies: samsung and micron. memory kingpin samsung alone controls one third of total dram supply. in november 2002, infineon and nanya formed a joint venture to help each partner expand its position in the dram market while sharing development costs. and in march this year, elpida and powerchip semiconductor corp. signed a sales and purchasing contracts agreement for their own dram strategic alliance. <col width=45> <col width=43> <col width=67> <col width=67> <col width=67> <col width=55> <col width=67> table 1: worldwide dram market share ranking 2002 sales rank 2001 sales rank company 2002 dram sales $’ mill. 2001 dram sales $’ mill. % change sales yr-over-yr 2002 market share 1 1 samsung $4,985 $3,205 55% 32.5% 2 2 micron 2,794 2,324 20% 18.2% 3 4 infineon 1,965 1,175 67% 12.8% 4 3 hynix 1,962 1,768 11% 12.8% 5 7 nanya 844 333 153% 5.5% 6 5 elpida 615 874 -30% 4.0% 7 11 winbond 478 130 268% 3.1% 8 8 mitsubishi 362 261 39% 2.4% 9 10 mosel vit. 302 205 47% 2.0% 10 6 toshiba 287 459 -37% 1.9% 11 13 powerchip 260 95 174% 1.7% source: isuppli, february 28, 2003 <col width=50> <col width=51> <col width=70> <col width=75> <col width=59> table 2: top five dram suppliers' ranking q1 2003 q1 2003 sales rank company q1 2003 dram sales $’mill. q1 2003 dram market share q-to-q share change 1 samsung $1,112 31.1% -3.8% 2 micron $701 19.6% 1.8% 3 infineon* $613 17.1% 4.2% 4 hynix $444 12.4% -0.1% 5 nanya $156 4.4% -1.5% *infineon revenue includes dram licensing fees source: isuppli corp. may 2003 some industry observers believe the major memory suppliers deliberately and routinely conspire to fix prices in order to 1) influence earnings results or 2) potentially squeeze out some of the weaker players from asia. from time to time, these large suppliers withhold supply of dram and other memory components, thereby driving up prices and creating artificial demand that in turn helps earnings results. regarding the latter “conspiracy”, the three main culprits appear to be samsung and hynix, both of korea; and taiwan’s nanya. other observers argue that such structured, formal conspiring seems unlikely and that the price trends are seasonal only. so perhaps it is merely coincidence that samsung’s, infineon’s and, to some extent, micron’s 2002 results benefited from firming dram prices in the second half of the year – particularly considering the otherwise flat pc marketplace. the legal issue last year, the antitrust division of the united states department of justice (doj) in california launched an investigation into alleged “anticompetitive practices” such as price fixing in the computer memory markets, possible collusion in dram pricing, and manipulation of manufacturing capacity. subpoenas were served on samsung, micron, infineon, hynix, and a handful of other smaller suppliers. at the time, micron was preparing to pay around $3.4 billion to acquire hynix, but the deal fell through – coincident with the doj investigation. speculation at the time was that the doj was looking into collusion on memory chip pricing, although industry participants and observers seem divided on whether the big chipmakers have ever engaged in such practices. however, at an industry conference in april 2002, dell computer's chairman and ceo, michael dell, commented that memory suppliers must be benefiting from the recent rise in memory component prices. these companies supply dram to pc-makers like dell for use as the main memory in desktop and notebook computers. "there was some cartel-like behavior by a number of dram suppliers," dell said at the conference. "there was an assumption by some of the companies that they could have both an incredible increase in the price of dram and [increased] demand at the same time. the world just doesn't work that way." i understand the exact nature of the doj investigation has not been disclosed. many have speculated that volatile price patterns in the memory market, coupled with the “coincidental” unison with which dram prices (from all the large suppliers) react, suggests that suppliers may have acted to adjust prices accordingly. these patterns may have raised the issue of possible collusion. however, the doj did confirm that its antitrust division was conducting an industry-wide investigation. the investigation reportedly revolved around an alleged effort among dram suppliers to influence, or “fix”, prices. related legal actions alleged that price increases by the major suppliers of memory components such as dram, ddr and sdram had violated california antitrust laws and forced buyers of dram to overpay for the chips purchased during certain times when dram prices were inflated. price fixing: a conspiracy or long-standing competitive dynamic? it is hard to imagine such fierce competitors as micron, samsung, hynix, and infineon acting together. some believe that these companies have been illegally cooperating in order to salvage profits in what has become a cutthroat business with razor-thin margins. however, it is often the case that competitive dynamics, such as price fixing behavior, particularly in a heavily concentrated industry, do not need to be part of a formalized process. these suppliers have all been observing each other’s moves for decades. as a result, they all have extensive competitive intelligence “institutionalized” within their respective organizations and management ranks. competitor information can be readily gathered since they have all shared, and continue to share, customer, supplier, and vendor relationships. speaking the language of “the nod, the wink, and the whisper” these players recognize all the signals and know how to signal each other. consequently, these suppliers can readily “read” each other’s market signals, infer conclusions about each other’s predictable behavior, and react accordingly. some industry analysts believe there is no hard evidence that dram suppliers have ever fixed component prices. there appears to be no direct evidence from the suppliers, or the pc manufacturers, and collusion appears less likely during periods when memory prices were particularly low. that being said, after a bad year in 2001, the memory suppliers enjoyed a solid recovery in the first quarter of 2002. the average selling price for a 128-megabit sdram jumped over 120% from around $1.70 in the fourth quarter of 2001, to around $3.75 in the first quarter of 2002. after a seasonal easing of dram pricing back to $2.50 last summer, bad for memory suppliers, in the fall ddr was back trading at a nice premium, and advantageous for the memory producers as they ramped capacity.  impact of dram prices in december 2001, contract prices for dram were less than $1. a slump in demand for the standard 128-megabit dram chips had sent chip prices plummeting and this, in turn, negatively affected memory chip economics and results at the major chip suppliers. however, by the end of may 2002, dram prices were moving back up into the $4 - $4.50 range, and industry chatter in asia at the time suggested price fixing maneuvers among the major dram suppliers. possibly it was the dramatic decline in dram prices in 2001, followed by the steep price hike in 2002, which set off the doj investigation. memory suppliers were trying to recover some economics after a period of selling units significantly below cost. the slump in demand for the standard 128-megabit sdram chips sent prices plummeting, and adversely affected all the major players. many industry observers have noted that the dram industry has been subject to wild price swings over the years. the synchronization of pricing trends seems to be easily explained. while dram suppliers find out from their customers what they are willing to pay, it's entirely possible that some suppliers have other channels to obtain each other's pricing information. it is a fiercely competitive market but cooperation, particularly in support of higher prices, would benefit all of the suppliers. memory component pricing can have a dramatic impact on pc manufacturers such as dell, since main memory can make up 5-6% of a pc's total materials cost. easy to see that major pc manufacturers may have felt that, during times of steep increases in dram pricing, they were being hit by a unified action and suspected dram suppliers of colluding to “fix” prices higher. on the flip-side, from the dram suppliers’ perspective they went from losing a substantial amount of money in 2001 to trying to make up some of their prior year losses in the first quarter of 2002. spot versus contract prices as for the relationship between spot and contract prices, the dram contract price has recently moved up, but spot will always lead contract, at least in a normal environment. in spring 2002, there was a point where spot was below contract, which was a signal that prices were headed lower. last fall, however, the bias was up. and today, that bias is also up, after soft dram prices in the first quarter. until such time as the bias is down, if spot converges to contract by declining it will be a bearish indicator until proven otherwise. in march this year sdram was being priced above ddr in some lots, which was not a great sign. spot pricing on dram recently crossed above contract, after a period in the first quarter in which dram prices were falling. this crossover and upward bias in dram prices should provide some support and potential for more orders from dram companies later in the year. and this in turn should provide support for, and probably even propel semiconductor and semi cap equipment stock prices over the next three to six months. according to dram-exchange, dram contract prices have remained stable recently despite falling spot prices, and this looks set to continue through may. despite weak pc demand, hardware manufacturers continue to source dram from the contract market, which has supported memory prices. however, absent evidence of sell-through to end-users or an up-tick in demand, there is mounting concern that this will lead to channel inventory build-up and softening dram prices. dram-exchange goes on to explain that although the major dram suppliers are trying to keep unit prices above $3, it is possible that this price point is not sustainable beyond mid-may. consequently, as dram suppliers are pressured to ship before the end of may, and as demand continues to weaken, this suggests dram price erosion, rather than ongoing stability. market impact of dram price “fix” one hypothesis is that the move up in the broader stock market in october-november 2002 was led by the semiconductor and semiconductor capital equipment [semi cap] stocks, which led all the tech stocks up, which in turn brought the overall averages up. but what was causing this upwardly mobile market daisy chain? the semis were quite possibly led up by the move in ddr prices. coming into late second quarter/early third quarter, there should be the usual uptick in pc demand as corporations spend out their it budgets for the year, which in turn should stimulate dram pricing. this happens every year; it is regular and predictable, and is irrespective of whether there is great pc growth, or just ok growth, or even no growth in the second half. so the stock prices of companies like micron could be expected to rise as dram prices firm. a common misconception in the analysis of the semiconductor business is that a shortage in ddr memory requires a substantial increase in semi cap equipment spending in order to boost ddr manufacturing capacity. in reality, ddr just substitutes for sdr [single data rate], so the only expense incurred by the chipmakers is for new masks (a kind of mould) for the different chip designs. this involves little more than a slight tweaking of the manufacturing dials. however, many investors believe that a shift to ddr on the production line requires new factory equipment, which is just plain wrong. still, this widely held belief in the investment community was one of the reasons that the fall 2002 spike in semiconductor stock prices was accompanied by a spike in stock prices of semiconductor capital equipment companies (such as kla-tencor and applied materials). in summary, history points to, at the very least, some coincidental factors leading to periodic spikes in ddr prices, such as that which occurred last fall. such factors include a faster-than-anticipated shift to ddr over sdr by pc-makers, at the same time that the seasonal third and fourth quarter inventory build was occurring. coincidentally, the dram situation in fall 2002 was similar to what happened in 2001. dram shot up, followed by the semis and the broader market, and then down dram prices came again, followed by the market. once sufficient capacity was switched over from sdr to ddr, and the seasonal inventory build had begun to slow, ddr prices began to fall – as would be expected - beginning in late 2002 and continuing well into the first quarter of 2003. anatomy of dram pricing ddr price increases could be considered to behave much like a soufflé – rising and falling on hot air. that is not to suggest that memory chip prices are not influenced by fluctuating levels of real pc demand; of course they are. but there is more to dram price patterns than ebb and flow of natural supply and demand. to date, there seems to be little, if any, hard evidence of collusion. still, the unison with which the large dram suppliers manage their pricing reactions is quite possibly more than coincidence. ---------- melanie hollands has over a decade of experience covering the technology and telecommunications sectors, from positions held in business strategy (mckinsey & co., bain & co.), corporate finance (salomon smith barney) and fundamental equity research (merrill lynch). while at merrill, she received an institutional investor all-star honorable mention for her cover of the pc hardware and wireless stocks. she covers: pc/server hardware, storage, enterprise software and operating systems, wireless hardware and software, data networking, optics, semiconductors, semi capital equipment, and telecom equipment. she is currently president of koala capital, which focuses on trading/investing in technology stocks, and from time-to-time undertakes (on retainer) technology strategy and fundamental research projects. editor's note: the opinions and conclusions in this article are solely those of its author and may or may not be shared by osdn editors and management.  slashdot it!   -    del.icio.us   -   digg this! related links sponsored links: comments on dram pricing: the fix is in note: comments are owned by the poster. we are not responsible for their content. title author date-time how about pointing out another fraud? anonymous coward may 12, 2003 04:10 pm like the fraud of motherboard manufacturers who list the capabilities of a motherboard as supporting 4 gb of ram in 4 slots, then finding out, well...it's really only 2 gb of ram, but all 4 slots can't be used unless you use ecc ram...(which is used, what,<nobr> <wbr></nobr>.05% of the desktop market?), and ohhh, btw, after testing, you only can use two (shuttle) or three (some of the others) of the four slots, without the full gb of memory (non-ecc) in the working slots...this isn't fraud? shuttle and the other motherboard manufacturers aren't defrauding the public by marketing motherboards that are marketed as being capable of being fitted with 4 gb of ram in four slots, when only two or three of the slots work (1st and third only of one or several of shuttle's amd boards, according to motherboard review sites' testing), and the slots can't take a full gb per slot unless ecc ram is used? that comes out to 2x512 mb of ram, for a grand total of 1 gb of ram maximum, unless ecc ram is used, on a board rated for 4 gb of ram. or for other manufacturers, where three of the four work, limiting to 3x512 or 1.5 gb of memory, on a 4 gb rated board.let's see if any of the motherboard review sites have the intestinal fortitude to make this charge. i've seen the memory issue buried here and there, but the charge has never been levied as far as i'm aware of.maybe not a big deal when the boards first came out due to ram pricing, but with price drops now, the fraud is the only limiting factor. # not fraud but poor qa anonymous coward may 12, 2003 07:48 pm most board manufacturers simply do not have the time nor access to the memory parts to test things. in order to "accurately" test the boards then they'd have to get modules from many of the memory manufacturers. this is unlikely. instead, they get what they can, test it, and then determine that it does or does not work. over the last few years, one module from one manufacturer will work and another module from another one will not work. is that the board or the module? neither being fraud -- just poor qa. # re:not fraud but poor qa ickusslime may 12, 2003 08:52 pm so what your saying is... if i have a poor qa department i can tell people one thing and actually give them another... thats called bait and switch and is considered fraud and is illegal. # not poor qa. fraud. anonymous coward may 12, 2003 08:58 pm there are many examples of motherboards which cannot work with all of the simm slots filled -- regardless of the manufacturer of the ram. i could understand how they might have problems with certain brands, mixing brands, timing issues with some simms, etc. i understand that the motherboard manufacturers cannot test every different simm in every slot in every combination.but, if they put four simm slots on the motherboard, they should be able to specify at least one fully-loaded ram configuration with which the board will work. if they say that the board works with ddr-333 ram and it supports up to 4gb of ram, then it should work with 4gb of ddr-333 ram. # i understand why you feel cheated anonymous coward may 13, 2003 12:14 am the motherboard manufacturers really should be more up front about this. however, the limits are real physical limits and not, as some suggested, necessarily showing bad qa.the problem is not ecc vs. non-ecc, it is registered vs. non-registered. registered dram has a buffer that makes it less of a load on the memory bus. unfortunately the buffering done by registered dram makes it slower while being more expensive to produce. only the professional market wants to pay that extra price, and they do not want memory contents to change randomly on them, so they will go for ecc every time. hence no registered dram without ecc, there is no market.high-speed busses are fickle things. i am constantly amazed that they work at all. # what are the consumers' options? anonymous coward may 12, 2003 08:54 pm what brand of dram should i go for, then? one that's actually trying to compete in price as well as quality? # re:what are the consumers' options? anonymous coward may 13, 2003 09:45 am yes<nobr> <wbr></nobr>:) # artificial supply shortages. anonymous coward may 12, 2003 09:37 pm it has also been suggested that some memory suppliers periodically withhold supply of dram and other memory components. this creates an artificial supply shortage, which in turn generates a level of artificial demand that drives up prices.hmm. sounds like what we live with every day with opec. # dram opec anonymous coward may 12, 2003 10:03 pm i've always wondered why the major memory chip producers did not create an opec type consortium for dram. they would be able to control the price of ram chips and hopefully hold it at a level that would be cheap enough to ensure brisk sales, while ensuring that they would make enough profit to a) keep their workers employeed all of the time, b) keep their production lines running at a certain capacity, c) be able to invest in memory chip technology.the dram constortium could raise the prices on memory chips to a point where consumers would find it too expensive to buy chips, but a) the smaller manufacturers could offer cheaper products b) like opefc the consortium does not want to alienate its consumers through higher prices.on another note, "regarding the latter “conspiracy”, the three main culprits appear to be samsung and hynix, both of korea; and taiwan’s nanya." though these three companies are geographically more closely located than the other major companies, it does not necessarily mean that they would want to price fix. # shhh. sloveless may 14, 2003 11:07 pm don't give them any ideas. # hemorrhaging anonymous coward may 12, 2003 10:21 pm the dram industry is hemorrhaging like never before. take a look at micron's last quarter. they lost more than half a billion in 3 months. there is no doubt that if there is price fixing going on micron is not benefitting from it.here's a great source to see the trends in dram pricing.http://www.converge.com/ewebapp/jsp/pricetrends/c<nobr>a<wbr></nobr> tegorydescdetail.jspit shows no indication that price fixing is currently occuring. the only products line that appear to be surging in price are the pc100 and pc133. this is very common at the end of a product's life.i don't like the inference this article gives. it leads the reader to believe that the dram manufacturers are getting fat using price fixing tactics. i think that it's closer to the truth that dram companies are with-holding supply in order to drive the price up so that they can survive.in micron's case, they are loosing almost 1 dollar to for every dollar they make in revenue. if the market does not see a turn-around soon, we could see the massive consolidation. this would lead to another industry similar to the cpu market where one manufacturer dominates. this will be infinitely worse for the consumer. # re:hemorrhaging cornstalk may 12, 2003 10:55 pm you say:"i don't like the inference this article gives. it leads the reader to believe that the dram manufacturers are getting fat using price fixing tactics. i think that it's closer to the truth that dram companies are with-holding supply in order to drive the price up so that they can survive."perhaps in the popular imagination these two things are somehow different, but speaking as a professional economist, i can assure you that they are one in the same. if a strategy of withholding supply actually succeeds in driving up prices, this can only be the result of some form of collusion among "competitors," -- however tacit it may be. a dairy farmer cannot drive up the price of milk by withholding his supply. only collusion among dairy farmers can result in a sufficient quantity being withheld to affect the price of milk. similar principles would apply in this market if it were a workably competitive one.i assume that the article is correct in its claim that a single supplier accounts for 32% of the product. in that case, he probably could drive up the price once or twice, for short periods, by cutting back his supply. but if the market were workably competitive, his competitors would learn take advantage of his withholding strategy, cut prices and sieze more of the market for themselves. as his competitor, for example, i might build up small surplus of chips that i could sell into the market the next time he staged one of his withholding ploys.it appears that this sort of thing is not happening, which gives plausibility to the claim that some form of collusion is taking place. that is perhaps not surprising, given that a 20% has been considered a rule of thumb maximum market share for any firm, necessary to ensure workable competition. # re:hemorrhaging anonymous coward may 12, 2003 11:19 pm agree that its about survival, rather than getting fat and rich. disagree that article presents that semi companies are getting fat on the "potential" practice of fixing. consolidation in the business, which is "commodity-like", and reference to "razor thin" margins implies declining economics in a susvival game. # ???collusion warpengi may 12, 2003 11:07 pm dram costs significantly less now than it did 3yrs ago. i remember jumping on a great deal in the fall of 2000 when i bought 128 mb dimms for $1/mb.it is hard for me to understand how there could be collusion and price fixing when the market price is so low.frequently over the past 2 yrs. the spot price droppped below $3/chip, the break even point for the manufacturers. if there is collusion it is resulting in significant losses for all the manufacturers.it seems to me that if i were colluding i would try to make it profitable. maybe i don't understand the ways of big business and that's why i'm not a ceo. # re:???collusion anonymous coward may 13, 2003 01:28 am this woman obviously has better information than the rest of us, but collusion is not as easy to pull off as it might seem (just ask opec, and their cartel's in the open). the incentives for cheating, and grey market behavior, are very high. if there were only 2-3 suppliers and they were all located in the same country, it would be different. # buy low sell high doctor digital may 13, 2003 09:52 am dram (ddr or otherwise) is small potatos in the fraud game.gas, tobacco, condoms, cell phone minutes, pick your ripoff.american consumers are the dumbest on the planet. we are consistently gullible and our government, corporations, and fellow citizens prey on us constantly.somewhere during my life span (celebrating 50 years of activism this year) things have changed a lot.many moons ago, businesspeople used to sit down with pencil and paper to determine how much it actually cost them to provide a product or service and would then add a profit margin to allow their business to grow and survive hard times. if you made widgets that cost some amount for raw materials and labor to produce, you added 10, 20, or 30% to allow for research or market swings.i am not sure when, but now the market is whatever some dumb ass is willing to pay. talk about inflationary !! that is the rule now. i have a friend that i argue about this with all the time. his opinion is that if consumers are willing to pay it, it's ok. this is against all my moral principles. i could not sleep at night if i did that.the moral: shop around for everything. get educated about what you are buying. other than used cars, the two best examples of things that most people just buy without "bargaining" are airline tickets and car insurance. get a quote and tell them you found it elsewhere cheaper. you would be amazed how fast they find you a "special" that you magically qualify for right away.suspect everything, question everything, and when in doubt, tell them the price seems too high. buy when prices are low. if i could store 1000 gallons of gas at my house during the winter to wait the bastards out during the summer, you bet i would.bottom line is that in this country, we can't even differentiate between what we "need" and what we "want".use your old computer until a new piece of software "requires" you to upgrade. do you need a 3 ghz machine with a 40gb hard drive and 64mb video card to do e-mail or word processing? hell no!!.think !!blood pressure coming down now... aaaaahhhhhh !!doctor digital # horrible horrible anonymous coward may 13, 2003 10:18 am its horrible how bad news is getting these days.i'm no fan of corporations, and no fan of price fixing but presenting this story with such a huge bent is just plain bad journalism.sensationalizing to draw readers in is a short term approach. writing for the long term takes facts like yours and remains impartial letting the reader draw their own conclusions.you write:i understand the exact nature of the doj investigation has not been disclosed. andto date, there seems to be little, if any, hard evidence of collusion. still, the unison with which the large dram suppliers manage their pricing reactions is quite possibly more than coincidence. how can you justify your title? "the fix is in" if there has been no conclusion by you or the doj about the culpability of the memory manufacturers.while the disclaimer at the bottom of your article says its an opinion...newsforge needs to focus on keeping opinion pieces in a separate section or use a different color. # linux??? dram??? anonymous coward may 13, 2003 12:46 pm i thought the articles in here are only around linux... # the first week of the ftc trial of rambus.... anonymous coward may 15, 2003 01:19 am here is some information you might find enlightening.... from the ftc trial...this case was originally about rambus committing fraud on the jedec sso by taking notes and later patenting the technology that was discussed in the sso, without revealing their ip postion to the other members, it had morphed to rambus violating some unwritten rules while a member of jedec, the memory industry's sso... now of course, the ftc has morphed that again, to a case of rambus not showing "good faith" to other members in the sso...i guess the ftc had to do something to save face after the cafc ruled that rambus did not commit fraud as previously alleged...now in the courtroom, the ftc case against rambus has taken some strange twists in only it's first week...the first ftc witness, from ami2/synclink/sldram and an official of jedec, testifies that jedec members understood that all ip interests must be disclosed during any discussions of any technology that was related to the ip. basically saying that every sso member gives up all rights to anything that's even vaguely discussed at the sso meetings..but, under cross examination by rambus attorneys, admits to patenting (claims he didn't know his name was on the patent, but did sign the affidavit as an inventor) technology as part of the synclink group, then presenting it to the sso for standardization without revealing that the technology had a pending patent with his name on it. and after first denying that he gave assurances to the sso that it would be royalty free, he later admits that indeed his presentation did say that (after he was shown one of the slides from the presentation that had the words "royalty free" on them). but, the witness, still under cross examination, goes on to say that "royalty free" did not mean free, but actually meant rand... ???now, this goes one step beyond even what rambus has been accused of... everyone is unanimous that ip disclosure had to occur prior to presenting any technology for standardization... that's about the only rule that was written down...the second ftc witness from micron technology first testifies that jedec members understood that all ip interests must be disclosed as early in the process as they are known.but, when cross examined by rambus attorneys, admits to taking notes on what was being discussed at the sso, going back to work and patenting those technologies. the same person presented that same technology for standardization, without any ip disclosure. one year after that technology was standardized, revealed his employers ip position in the technology and gave assurances to the sso that "there would be no problem with the ip " , still later admits that his employer sued another member company of the sso for patent infringement over that same ip. and still later admitted that they probably should have disclosed the ip to the sso at the time of presentation.the final, and probably the most important, thing that has happened is that the sso rep from nec, under questioning from the ftc, claims there was no ddr (a standard) technology in the rambus europeon patent application (that rambus filed prior to joining the sso). this european application became public (under the 18 month disclosure rule in effect) during the first year of rambus tenure in the sso and was basically disclosed when rambus disclosed the specification of their 703 patent, which on it's face showed both divisional applications and a continuation in part application... this european patent app was discussed within the sso, but was dismissed as not being issuable as a patent due to prior art.upon crosss examination by rambus attorneys, the witness finally admitted that the drawings included with that application did indeed show ddr technology (that rambus has been accussed of stealing from the sso)..all in all, so far the ftc has succeeded in making rambus look like one of the most ethical companies in the sso...the above information was distilled from my subscription to fredhager.com who has 2 reporters covering the ftc trial...you just can"t make this stuff up !!! # post a new comment name : anonymous title: comment: are you human? copy the word from the image on the left into the input box.   © copyright 2007 - sourceforge, inc., all rights reserved about linux.com  -  privacy statement  -  terms of use  -  advertise  -  trademark  -  ask linux questions  -  contact us  -  write for us  -  rss feed thinkgeek  -  slashdot  -  it manager's journal  -  sourceforge.net  -  freshmeat  -  surveys  -  jobs  -  pricegrabber

Acceuil

suivante

linux.com :: dram pricing: the fix is in  Teardrop Parts  How to fix the GST mistake  fix - Synonyms from Thesaurus.com  'Fix' is finished :: CHICAGO SUN-TIMES :: Robert Feder  EAZ-FIX, Instantly undo your PC problems  SIGN FIX sur SOCIETE.COM  Aural Fix Communique Online  Deeper level of interest in climate fix csmonitor.com  Fix APO ? - forum Snowboard  fix tps  Fix Blogger Sitemap XML Warnings Inside Google Webmaster at ...  :: SUSHI FIX RESTAURANTS :: WELCOME  Credit Repair by Credit Clean Help Repair Bad Debts  Encyclopédie méthodique, ou par ordre de matières - Résultats Google Recherche de Livres  NT4.0 too flawed to fix – official The Register  NASA delays spacewalk to fix solar wing International Reuters  : : RegistrySmart : :  Clean PC Software :: Registry Fix It!, Error Scan and Fix, Free ...  Eric's Archived Thoughts: Framework Fix  Futurist: To fix education, think Web 2.0 Tech News on ZDNet  Traduction Garbage Fix Me Now lyrics - musique traduite  Video: Your daily Tesla fix! - AutoblogGreen  Patch: fix for PR java.lang/339  Tom Tromey - Patch: FYI: Fix for PR 4859  Multi-Fix N.V. - Professional Signmaking Films signmaking vinyl ...  Flickr: FIX MY PIC  The Fix (Heroes) - Wikipedia, the free encyclopedia  Daily Funny Fix / Media  Dent Fix Equipment  HTML FIX IT.COM: A site with free scripts, advanced scripts ...  Journal de physique, de chimie, d'histoire naturelle et des arts ... - Résultats Google Recherche de Livres  Gaufrey: Chanson de Geste - Résultats Google Recherche de Livres  [Profil de fix] OverBlog - Le blog des blogs  [Profil de Fix] OverBlog - Le blog des blogs  Contes de Guillaume Vadé - Résultats Google Recherche de Livres  Sur le compte rendu au Roi en 1781: Nouveaux éclaircissemens - Résultats Google Recherche de Livres  Renaus de Montauban: oder Die Haimonskinder, altfranzösisches Gedicht - Résultats Google Recherche de Livres  Fixsuche.com Fix gesucht und schnell gefunden! Startseite  Dictionnaire historique de la ville de Paris et de ses environs - Résultats Google Recherche de Livres  Les vies des hommes illustres de la France, depuis le commencement ... - Résultats Google Recherche de Livres  Mercure de France - Résultats Google Recherche de Livres  Nukemods.com - FIX to the 5 last posts block on the front page  Nukemods.com - [FIX] Wrong url after search for "redirect"  Amazon.com: Fix-It and Forget-It Cookbook: Feasting with Your Slow ...  www.live.com fix - User JavaScript - Opera Community  Nvidia.com fix no longer needed - UserJS.org Bug reports - Opera ...  Registry Fix Free Download at CleanSofts.com - Fix Errors, Speed ...  System Mechanic Free Download at CleanSofts.com - Fix errors ...  Erotik-Shop.com - Fix it  Detalii si Statistici pentru http://www.fix-hdd-bad-sectors.com ...  Fairuse4wma fix dans la section BestOf  Bondage Fix: get your bondage bdsm fix here!  Titfix.com - Get Your Daily Fix of Big Tits!