Slashdot Mirror
FTC Dismisses Complaint Against Rambus
swordboy writes "A federal judge just threw out the FTC lawsuit against Rambus. This has been discussed at length here before but this changes the landscape yet again. An interesting, possibly coincidental item is that Intel just today announced a new and very powerful DRAM interface that bypasses Rambus IP altogether."
I fried a mobo on my Rambus system a few weeks ago and quickly found out how rare these mobos still are- and how little the price of memory has dropped over the last two years.
Memory is like disk space. The general population demands quantity over speed or quality. Rambus was a technology that never really trickled down to the average desktop.
Those that suggest you "dance like no one is watching" really want to see you make a complete fool of yourself.
RAMBUS is another company that is dedicated to making its money now through lawyers. Intel thought that they could take more control of PC design my picking a patented memory structure, and RAMBUS was the perfect lackey ito accomplish this. Their contract with RAMBUS would have had RAMBUS paying Intel back once RDRAM sales exceeded a certain amount. It was a win-win for those two companies, and lose-lose for everyone else due to higher long term prices for all users, and manufacturers.
The reason for this is the RDRAM design. It takes more space on a wafer to produce, and that is why it costs more ( commission to to RAMBUS is another part, but the size difference is the key cost difference ). So memory prices would have been much higher, and Intel would have been able to squeeze AMD more due to the patented bus that RDRAM uses.
If you go back in time, it was exactly as Intel was about to force RDRAM down everyones throats, that AMD released the Athlon. Suddenly there was an alternative to Intel in performance, and by not using RDRAM, the price difference was extreme. This is the point that AMD surged ahead in market share, and while the inroads they made were overall not significant, they were enough to show that not everyone would be pushed around.
RAMBUS did come up with some interesting design innovations, but as soon as the writing on the wall was that RDRAM was dead due to lower prices with DDR, they turned into SCO by suing everyone that was making DDR, by use of info they had taken from JEDEC and adding it after the fact to pending patents from RDRAM. Another stellar example of USPTO excellence. RAMBUS is dead, but someone wants to make money from the rotting corpse. Just compare how similar the lawyers fees are for RAMBUS and SCO.
the concept is indeed pretty cool, though you'll need some tough lil drivers that can handle incoming voltage swings while it's driving.
No you don't. You already need to drive a line that's got a charge on it from the stuff you previously drove onto it. This doesn't change that. The local end just sees the far end as being terminated by a resistor to a voltage that is either low or high, rather than being terminated by a resistor to a constant voltage.
Driving both ways simultaneously, though, is very cute.
The downside is the need to daisy-chain. That means you're driving multiple lines at 3.6 Gbps on EVERY chip, ALL THE TIME. That's a LOT of power. Even if you interrupt the daisy chain at the selected chip (and arrange things so that the quiescent states of the transmitters at both ends of an idle line match) it's still a lot of power unless you localize most of your memory access to the closest chip.
Bantam Dominique roosters crow a four-note song. Once you've heard it as "Happy BIRTHday" you can't NOT hear it that way