27 Apr
2003
27 Apr
'03
2:31 p.m.
Jeffrey Sharp wrote:
On Friday, April 25, 2003, Wayne M. Smith wrote:
That is cheap, uninformed, and baseless.
I worked at a company that used Rambus back in 1993, getting 600 Mb/s
per pin in the days before SDRAM even existed. At the time, many people
didn't believe it was possible. Remember that at that time routing 66
MHz signals on commodity PC boards was pushing the limit.
Working with it was certainly a challenge (testing the interface posed
special challenges), and the price per megabit of it never became
competitive with commodity DRAM.
Rambus really did have a vastly superior interface technology. The
physics of it were better, the architecture of it was better. It really
did work.
What happened with respect to the JEDEC standards meeting may be a
different story. But to claim that they set out to deceive people and
take credit for things they didn't do is contrary to fact. As for point
4, part of the reason Rambus has had a spotty history is they didn't do
enough of that.
In my experience, its [patent system] primary use
is as a shakedown device
to extract royalty payments. The idea is not to actually thwart
competition, but to threaten it enough that accused infringers, fearing an
injunction, capitulate and pay a royalty.
There is a neat little business process in use today by a few corps. I call
it the Rambus method:
1. Hire some lawyers
2. Get some illegitimate patents
3. Deceive lots of people
4. Profit!!!
I wonder if any company has tried to patent the process, and if so, if the
company's employees had any idea of the irony.