Rambus rambushes DDR camp
Weekly peregrinations from a Vulture's Nest
Column Surely the prayer of every Rambus suit as she or he rises in the morning must be "give us this day your daily bread"?
As the sun set over Mountain View last Friday, local residents must have been startled to hear hallelujahs and praise the lords from the usually silent Rambus campus as the suits changed into their cashual gear and boogied on down to the local burgher bar, drinking good Californian wines just before the warehouse burnt to the ground, with 90,000 gallons going to hell.
Well, the God of Stocks and Shares certainly shone on Rambus last Friday, aided and abetted by angels such as market maker Morgan Stanley Dean Witter and Toshiba, the semi-firm which likes to say yes. The IP firm, which split its stock four ways the day before, saw its new share price soar to $83.375, a rise of $26.6875 on the day. Will it make $500 before the end of the year, we were bound to ask ourself.
It was another of those oh-so-tasty Rambus sandwiches that we're beginning to think should be high on the list of toasties with the mosties for the hungry IT-eater. This time, instead of Tom Pabst being the morsel between two Rambus suits, as he was at Computex, the two halves of the bread were Rambus press releases and the meat was the share price.
At close of play Thursday, Rambus Ink announced that Toshiba had concluded that it ought to pay for making double data rate (DDR) memory and synchronous memory (SDRAM), as well as paying for the Rambus stuff too. Would this move splinter the Dramurai?
Friday dawned on Wall Street, and it was apparent that shareholders were getting very excitable indeed, with a buying splurge that resembled the rush which occurred when Intel's Dr Albert "even my mother knows about copper" Yu, told the world that Willamette was a Rambus only product.
At close of play Friday, the suits at Mountain View rambushed the world again by issuing a press release more or less saying, "look, Rambus owns DDR," here's a new technology we've invented based on DDR which is faster than our current technology and by the way we own DDR. (You can find the Friday release on the Rambus web site). Further, it even managed to roll out Toshiba again to apply seasoning to its QRSicL (quad rambus signaling (sic) level) announcement, heading the release: The World's Fastest Bus Technology.
(Young Master Magee knows only too well that the world needs faster buses because he was late to school Friday morning, the 183 from Harrow to Kingsbury having failed to provide enough bandwidth, but surely it takes the biscuit for Rambus to quickly shrug its old suit off and squeeze into the DDR T-shirt without so much as the faintest virginal blush?)
For the fact is, as Bill Gates told the world+dog only last week, these matters will be decided by a higher court. Hitachi and Rambus are currently engaged in bitter litigation. Rambus has patented DDR and SDRAM designs and that scrap will take some good time to resolve before the litigants drop exhausted and swivel-eyed, in a heap.
Meanwhile, common sense still works. As one geezer at Computex pointed out to last week, folk are usually very happy to pay more for technology which gives them very much faster speeds -- witness the rush for AMD 1GHz Athlons earlier in the year. But, on the other hand, they feel more than a little miffed if they're paying through the nose for RIMMs that demonstrably don't offer far superior performance.
Plucky little Taiwanese chipset vendor Via, which looks like it will have a huge chunk of biz by the year end and which, by the by, will soon list on US exchanges, just like foundry TSMC, must be terrified that it will have to turn over all those naff DDR-266 T-shirts it was distributing at Computex if Rambus pulls off this DDR stuff. (Via suits merely equivocated when we asked them about its relationship with Rambush).
And the Taiwanese motherboard vendors, by the way, still have the needle about those extra pins that Willamette 479 will sprout in Q2 next year when Tulloch has its day. Would it be beyond the realms of possibility that the additional pins in W-479 are there to support QRSicL? ®