Samsung Caves To Rambus Royalties

denominateur writes: "According to this story, Rambus now made a licensing agreement with memory market leader Samsung that will allow Rambus to collect royalties on virtually all the computer memory produced by the market leader. Who will stop Rambus from getting more and more money on patents that basically don't make ANY sense?" Well, evidently, Micron and the two other companies fighting the Rambus patents in court seem like possibilities. I wonder what happens to those royalty deals, though, if the company goes Rambust ...

Comments

RMBS is losing it's props

[kalinh]

This is actually a pretty old story, well old as in at least 12 hours. I've been following the company pretty closely cause I shorted their stock abour 3 months ago. They dropped down as low as $36 a share yesterday as investors finally took their collective heads out of their asses and realized that Rambus has quickly developed one of the worst brands in the industry. Today's great news only resulted in a relatively mild increase (8% or so).

The point is that the company has been artificially propped up in stature and prominence for a long time and these supports are slowly being removed one by one. But it's taking a long time.

Intel has gone sour with RMBS and according to the public portions of the contract they signed with them all they have to do to cancel their commitments to promoting RDRAM with their new systems is write a letter terminating the contract. Why hasn't this been done then you ask? There are portions of the contract that are blacked out and the speculation is that RMBS has Intel by the balls on some IP issue that no one is talking about. Though that's just a good theory at this point.

At any rate, even if RMBS manages to collect royalties on SDRAM and DDR-RAM and it still looks to me like Micron et al have a strong case on the whole RMBS breaking the JEDEC agreement and patenting jointly developed technology.

Intel put together a group called the Advanced DRAM Technology (ADT) to develop a future RAM standard slated for 2003 and they didn't invite RMBS to the table. Regardless of whether RMBS has any claim on the patents that cover SDRAM and DDR, it's obvious that the rest of the industry (regardless of Samsung's licensing deal they are also working on the RMBS free standard) will be working very hard to bypass RMBS royalties as soon as possible.

Even if RDRAM was viable on the desktop (it's not, as others have pointed out), it's gettign to the point where associating with RMBS is bad public relations. Doesn't it make you feel sick to think that your new PS2 purchase has contributed to the financial health of these leaches?

There is a great article about this at Tom's Hardware from back in July. It makes some long-range predictions, many of which are now being played out nicely. The whole article and, actually, all of the Rambus coverage on Tom's is excellent. Hey while I was there looking for the link I found this new bit about the Samsung deal which totally jives with what I just wrote and also points out that many analysts beleive RMBS gave Samsung major concessions to stage a well-timed deal and that Samsung is possibly avoiding legal costs while waiting for Micron et al to take RMBS to court. It's also woth noting that Samsung has been one of RMBSs biggest allies for a long time.

Do patents push Innovation?

[Veteran]

The answer to that question - as is the answer to most questions - is Yes and No. Nothing is all good or all bad. The universe works on Yin and Yang; something which is good has some bad points - something which is bad has some good points.

When I say that something is bad, I recognize that it has some good aspects. I'll give a couple of personal examples on how the lack of a patent stifled innovation.

Back in the early 1980's I made a trivial invention: a comb which wouldn't create static in your hair when you used it. It was a very simple invention; all you have to do is add a little conductive material such as graphite to the plastic in its molten state and the comb no longer will generate static.

I applied for a patent and took the invention to one of the major comb manufacturers in the states. They were very happy to see me and excited about the prospect of being able to get an exclusive license on the 'technology'; they could see that being able to advertise a "static free" comb would create a competitive advantage for them.

However, the patent office rejected my application (the interfering invention was a comb in a xerox copier - designed to drain off static charge from paper.) When this happened the comb manufacturer lost interest. What they said is that without a patent they would have no competitive advantage - as soon as they came out with their anti-static comb so would everyone else - so they weren't interested in producing the combs. While they might be an advantage to consumers they would be no advantage to the manufacturers without a patent, and to this day they don't exist in the marketplace.

The second example comes from a somewhat more advanced invention: a differential microphone which cancels out feedback problems in live music. The mic works by having to microphone elements wired in reverse phase - one closer to the mouth than the other. Feedback from a speaker is rejected as common mode noise while the voice comes through as a differential signal. NASA has this patented as an external noise cancellation microphone for space capsule use. However the same setup can be used for PA or live music and cancel out most feedback. The results are dramatic: minus feedback, microphone speaker combinations produce much less 'mechanical' sounds; the voice from the speaker sounds very lifelike.

I decided to forgo getting a patent; my earlier experience had cost me a lot of money and I wanted to do the right thing: I released the invention to a major microphone manufacturer - > placing it in the public domain . The results were the same as the comb: no patent - no manufacturer interest. Now 15 years later no one has seen a differential microphone on the market place because the manufacturers would not benefit from its existence - even though the public would.

I hope from these two examples that everyone will be able to see that as long as we live in a world with competing manufacturers of commercial products - that a lack of a patent can and does stifle effective innovation.

Lets all tell Rambus to Blow Us

[Greyfox]

Kind of under the Radar screen, QDR lurks. If it's not bound by Rambus patents, Rambus could be doing us all a favor by driving the prices of the older tech up to the point where QDR gets adopted much more quickly. Although the Q and A page doesn't specifically mention Rambus, it does say that all three parties involved in the creation of QDR collectively own the patents on it. Those parties do not include Rambus.

I'd find it most amusing if Rambus ends up forcing its only worthwhile property (Its patents) into premature obscelescence through its collective greed and stupidity. I wonder if the company leaders think they'll be able to get jobs with any other tech company given their highly visible performance in driving this one into the rocks.

Re:Lets all tell Rambus to Blow Us

[milkman1]

Some how I just don't see people buying SRAM (QDR is SRAM) of any kind for main system ram anytime soon. To make something like that pay off, Rambus would need to charge royalties of about $2 a megabyte. SRAM needs 4 transistors per memory cell, while DRAM (including SDRAM RDRAM etc) needs only 1 transistor and a capacitor. Thus DRAM is about 4 times as dense and 1/4 the price of SRAM. In modern applications, SRAM is only used for very high speed applications (L1,2,3 caches for CPUs, Supercomputers, high speed networking hardware) and for very small (8kbyte) amounts of memory where the cost of the extra transistors in the circuit is offset by not needing circuitry for refreshing the memory (smart cards, microcontrollers) and low power devices (graphing calculators) (SRAM which is not being accessed disipates almost no power, while DRAM needs to be rewritten constally) In general DRAM is the memory of choice for PCs. In addition to main memory, it is used on video cards (SGRAM, VRAM, WRAM, DDRSDRAM etc are all DRAM), hard disk drive, hdd controllers, cdrom drives etc. SRam is typically found only in the L1 and L2 caches.

Patent system problem?

[atlep]

Rambus CEO Geoff Tate: "We think it would be difficult, if not impossible, to develop a competing technology to RDRAM and not infringe on our patents," he said. "We are extremely confident in our legal position."

Patents exists so that those who spend resources on research should be able to get something back, and thus stimulate more research.

What this shows is that the way the patent system works, it limits competition and makes research less interesting. Even if someone can come up woth some better technology, the economic advantage of doing so is partly removed.

Looks like the patent system may in, in some ways, be working against what it should be working for.

Unfortunately, Rambus will never die.

[AFCArchvile]

Rambus currently has a stake in PlayStation 2; the system RAM. Therefore, not only is Rambus sucking the blood out of legitimate RAM manufacturers through royalties, they're also being kept alive by the largest media juggernaut in the world, Sony.

Therefore, if you truly oppose Rambus, don't buy a PlayStation 2. That also holds true if you oppose SDMI, the DMCA, and the MPAA. The PlayStation 2 is the epitome of media corruption, and the corruption of the world by the media companies. By watching the facial expressions of people playing the PS2 demo unit, I learned the true meaning of "the idiot box".

In short, oppose Sony's monopolistic dreams: buy a Dreamcast.