Micron has filed patent-infringement lawsuits against several memory makers, including Micron Technologies, Infineon and Hynix Semiconductor, formerly Hyundai.Gee, these mudslingers can't even copy the script correctly!
By the way, here's a response to the article from a Rambus board member:http://boards.fool.com/Message.asp?mid=15352384What Spooner doesn't say is that the court appointed experts examined the patents and Micron's SDRAM specs for 5 months and determined that Micron infringes on the Rambus patents. The Monza judge denied the injunction but the infringement trial will continue in Milan.
I feel sorry for Mr. Spooner...I'm afraid he will be very busy over the next several weeks reading his email....and dining on crow.....hoemee
I think Spooner needs an editor or proofreader. From the article...Micron has filed patent-infringement lawsuits against several memory makers, including Micron Technologies, Infineon and Hynix Semiconductor, formerly Hyundai.Shouldn't that read "Rambus has filed..." or is Micron suing itself?
Shouldn't that read "Rambus has filed..." or is Micron suing itself?This may be a way for Micron to raise capital and offset the huge losses they are experiencing:)hoemee
DD,In your link, should it have read Rambus instead of Micron?Micron has filed patent-infringement lawsuits against several memory makers, including Micron Technologies, Infineon and Hynix Semiconductor, formerly Hyundai.Re Micron suing Micron Technology, or are they two different companies and/or does Micron have patents for which to sue?Confused.Harry
DD,Should have read the thread first.Harry
You all know my opinion on this matter and it has little to do with whether or not Rambus wins the litigation. RDRAM is superior, is creating superior solutions, and scales better than anything, including for those 4 ghz chips mentioned in the article. Your not gonna put out 4 ghz chips without a new memory standard.It is a bunch of B.S. to say the industry prefers "cheap memory even with the performance hit." Some how, despite this, the expensive SDRAM replaced EDO. And I still think it is questionable, not improbable, but very questionable whether DDR can scale, and if so, what sort of expenses in the total system cost will be required for it to do so.This in mind, the article does lay out the worse case for Rambus. A scenario, although unlikely, is still in the realm of reasonable probability. I think we need to watch two things:(1) How the 845 DDR chip does in 2002. Is there some sort of irrational desire for a mismatched P4 with DDR1600 out there?and(2) If DDR can be successfully brought to market, cost-effectively with system design and all, in faster chips. As an example can AMD use it in their 2 ghz plus chips (said chips still vaporware for AMD, not so for INTC).I think the odds are a lot higher than #1 will be more likely than #2, but we shall have to see how it goes. I think both 1 and 2 need to happen for DDR to be the threat described in the article.Thus, to conclude, game is not over. But there is no question RDRAM is in the lead and pulling away. DDR does have a path to victory, but it requires the above two factors to coincide, IMHO. So keep your eyes open and your eye on the prize.Tinker
I haven't listened to the conference call yet on Rambus but I have been reading excerpts and posts in regard. A few things are hitting my red alert button and I hope others who have listened to the CC can maybe address the issues.But first is in regard to the "flat fee" licensing arrangements that some customer renegotiated for the next 4 quarters. Rambus management has adamantly stated that the VA litigation had no effect on current licensees, yet this agreement, if the information was relayed to me properly, is based in part on successful resolution of the litigation. That things go back to the normal rates if litigation then resolved.Now, it is true that this could just be market heding by both sides with volatile DRAM prices as they are. But if memory servers the litigation tag was tossed on this agreement as well. If so, this bodes very, very, poorly on RMBS management. I will not own stock in a compamy that won't be forthright.The second big issue is what looks like an enormous increase in accounts receivables. RMBS never has anything but nominal receivables.I just haven't had the time to listen to the CC so any input would be welcome. But just as the strategic long-term situation looks promising for Rambus, the short-term story may be turning and I'd like to better assess what is going on here. The short-term may play a significant role in future long-term developments.Thanks all.Tinker
P.S.These developments were completely unexpected. I can't verify if the news as relayed to me is accurate, misleading or what. However, if accurate, and in context, then ouch, ouch, ouch!!!
You all know my opinion on this matter and it has little to do with whether or not Rambus wins the litigation. RDRAM is superior, is creating superior solutions, and scales better than anything, including for those 4 ghz chips mentioned in the article. Your not gonna put out 4 ghz chips without a new memory standard.Did you really miss out on nVidia's nForce? It's a dual DDR solution giving a total of 4.2 GB/s of bandwidth with superior latency. What can Rambus show for? 3.2 GB/s bandwidth with bad latency.Not scalable you say? Samsung is sampling their new PC2400 DDR. In a future version of nForce, that gives 4.8 GB/s bandwidth. We are also seeing QDR on the horizon. By the time dual DDR reaches the end of its line, dual QDR will bring the bandwidth up to 6.4 GB/s bandwidth and beyond.Onar.
Did you really miss out on nVidia's nForce? It's a dual DDR solution giving a total of 4.2 GB/s of bandwidth with superior latency. What can Rambus show for? 3.2 GB/s bandwidth with bad latency.Yes, and how many pins are involved? How many line traces on the board? How much noise and EMI? There are reasons that working with DDR is very difficult and why it is very difficult to scale.Yet another "announced" but not delivered product. BTW this nVidia device, appears to be more of a niche product for the high-end. Lets see some mass deliveries of the product and benchmarks thereon until we hoot our horns in regard. DDRs record to date is 0-3 in CPU products that deliver any measurable benefit. This one might but will be at the very high-end on expenses.So when I say lack of scaleability I mean how much noise, pin counts, etc., can designers design around. And the other thing you forget is that DDR's efficiency goes down the more and more bandwidth that is needed so when you quote peak bandwidth you are not giving a lot of useful information as sustainable bandwidth will be less than 1/2 of that.Tinker
Did you really miss out on nVidia's nForce? It's a dual DDR solution giving a total of 4.2 GB/s of bandwidth with superior latency. Didn't you mean "It will be a dual DDR ..." It isn't out yet, and if previous problems with 266MHz DDR are any indication nVidia will face a lot of problems making that solution work. Plus you don't mention that the 4.2 GB/s includes graphics bandwidth. When I see it for sale by major OEMs and see how it tests, then I'll give this proposed solution credit if it so deserves. It does look interesting, but I'm skeptical.The true test won't come until next year when we can compare DDR and RDRAM on more comparable systems. Until then I'm going with what Intel says: RDRAM is the preferred solution.Ben
I finally listened to the conference call, here are my impressions:http://leviticus.boards.fool.com/Message.asp?mid=15365907&post=trueNothing spectacularly insightful or anything. The boat's not broke, nor is there a leak or anything, just it is far from certain just how far from shore it'll travel. Gotta love boat analogies.Tinker
Best Of |
Favorites & Replies |
Start a New Board |
My Fool |