Rambus Claims It Was Price-Fixing Target

conq writes "BusinessWeek reports on the latest developments in the Rambus/Micron saga over pricefixing." From the article: "One e-mail, dated June 5, 2001, from Micron Vice-President Linda Turner to other Micron employees was in response to worries about prices on DDR-DRAM that had been falling. 'No problem!,' Turner wrote. 'We want DDR to explode in 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).'"

Comments

Rambus was overpriced and underperformed.

[Otis2222222]

I still remember quite vividly the feud between Tom's Hardware Guide and Rambus that resulted in Rambus' stock price tumbling. Even if the other manufacturers kept their prices artificially low (which I doubt), Rambus was easily twice as expensive if not more than the same amount of DDR would have been. The fact that real world performance tests didn't back up the hype that Rambus promised was the nail in the coffin. If it was priced lower, they might have had something. I thought the technology had potential, but it's absurdly high price kept it from ever being realized.

Or more likely...

[LordKazan]

... it was the fact that Rambus is SHIT

"PC800 RDRAM, which operated at 800 MHz and delivered 1600 MB/s of bandwidth over a 16 bit bus using a 184 pin RIMM form factor"

"Compared to other current standards, Rambus shows significantly increased latency, heat output, manufacturing complexity, and cost.[citation needed] PC800 RDRAM operated with a latency of 45ns, compared to only 7.5ns for PC133 SDRAM."

then squashed by

"DDR SDRAM, introduced in 2000, operated at an effective clockspeed of 266 MHz and delivered 2100 MB/s over a 64-bit bus using a 184 pin DIMM form factor."

not to mention needing CRIMMS or whatever they called the terminators

http://en.wikipedia.org/wiki/RDRAM
http://en.wikipedia.org/wiki/DDR_SDRAM

Collusion and rights

[Anonymous Coward]

Take a look at the prices of state-of-the-art SDRAM before and while RDRAM was on the market. Notice how 66Mhz 16- and 32MB SIMMs start out pretty expensive in the mid to late nineties, and are later replaced by equally expensive PC100 SDRAM. Then the price of PC100 SDRAM suddenly drops around 1999 to dirt cheap, and then the cheap PC100 is supplanted by equally cheap PC133, PC150, and PC200 SDRAM. Then just a short while later, really expensive DDR SDRAM with data rates in multiple gigabits per second hits the market and we've been paying for memory in a new higher price range than the old pre-RDRAM high price range ever since.

Now correlate that with the introduction and failure of RDRAM in the market and you'll see that PC100 prices dropped not long after it was introduced, and then fast, expensive DDR SDRAM came to market around the time RDRAM became irrelevant. Of course, that's just circumstancial. A lot of different market forces could have caused that kind of price movement.

RDRAM had a lot of technical problems with it. It did run hot, it did ride on funky slots, it was complex to manufacture, and for a variety of reasons, it cost a lot of money, not least because Rambus wanted to recoup the costs of developing the one advantage that RDRAM actually did have. That one advantage was that RDRAM was as fast as Intel's top-shelf CPUs. You could build a PC with a 400 Mhz Pentium II, a 400 Mhz FSB, a 400 Mhz north bridge, and a 400 Mhz memory bus leading straight into 400 Mhz memory, but only if that memory was RDRAM. Of course, 200 or 266 Mhz would have been just fine for most applications and even for most benchmarks. Matching speed with the CPU was overkill and Tom's Hardware knew it, among others.

The things that made RDRAM faster than contemporaneous commodity RAM were the patented designs of Rambus. Their problem was that they came to market seeking tech journo headlines at a time when the average PC consumer was fixated on the CPU speed, assuming that if they could match CPU speed with their RAM and get it written up, people would 1) stop fixating on CPU speed and 2) notice that commodity RAM wasn't cutting the mustard anymore.... and they overdid it, and overdoing it cost more than it needed to all the way down to retail.

Having said that, the evidence is only now coming to light that RDRAM wasn't killed by its own problems. It was killed by commodity RAM manufacturers flooding the market with cheap PC100 and PC133 RAM. So cheap that the cost curve of settling on the faster RDRAM part didn't make economic sense for most system integrators or their customers, despite the technical advantage. So RDRAM dies a quiet death of irrelevance around roughly 2002. Boo-hoo.

What happened next is the part that Rambus is currently seeking redress for. DDR SDRAM came to market, and we all know how it works and why it's exactly twice as fast as conventional SDRAM. What most people don't seem to understand is that RDRAM was DDR. That 400 Mhz RDRAM part actually used a 200 Mhz clock, and the FSB, north bridge, and memory bus of an RDRAM-capable motherboard were also DDR. Rambus developed DDR and holds the patent on it, among other things that have shown up in modern commodity RAM.

So let's recap. Rambus came to market with a problematic yet superior product which was ahead of its time in a market dominated by a few large manufacturers of commodity parts. The major manufacturers got in touch with each other to temporarily fix prices far too low to justify adoption of the problematic yet superior product which was ahead of its time. RDRAM became irrelevant, and the major manufacturers believed that Rambus had also become irrelevant. Once that happened they started using Rambus technology in their own products as the market needed it, while colluding to bump prices back up where they wanted them all along.

Since then, the post-RDRAM high price fixing has been proven in court. Rambus has kissed and made up with Infineon and Elpida with patent licenses and settl