Skeeter Bug,
intel engineers who saw rmbs as a joke are either fired or hated
I think that engineers' opinions are usually not that unified, but I think there was a broad spectrum of opinions and enough of the good ones were doubtful about the strength of the idea that their concerns should have been considered. This however, is not the first time (and not the last) where otherwise brilliant engineering executives look purely at the dollar signs in a business opportunity and chose to downplay/ignore the engineering risks or novelty of an idea.
And the outcome of this business decision could have come in either way. If RMBS was careful with their engagement in JEDEC, if there was no prior art w.r.t. the bus for rmbs ideas, if some patent was worded differently and applicable to ddr/sdr, if i820 had worked initially, PG and his friends who went for this could certainly emerged as heroes, AMD and other cpu manufacturers would be burned badly.
The world would have belonged to INTC to do as they please.
There for the grace of god...
Sometimes engineers and scientists like some here can get a bit extreme in their dislike and see news filtered in the way their prejudices are tuned. I actually read enough and heard enough to assume that the jury believed they amended their patent applications to match the jedec standard, going back to read it, I don't see any support for this viewpoint, so I am probably mistaken there.
But given what many techies in the trenches of techland are saying about rmbs, if someone said that rmbs management waved dead cats in a cemetery during full moons, some would probably believe it. Heck, I'd wonder if it is true about it for a few seconds too. Everyone I know don't like these guys.
Frankly, with the flipchip technology that removed pad-limitation in silicon, I just don't see the same long term value in quad-pumped pin-reduced designs anymore. Sure, in the short term, these forms of flipchip packaging are going to be premium, but the day when they also become commodity is not that far off. There is a small window when the pin reduction is valuable. After that, there is no clear advantage, given the tradeoff in complexity.
SbH |
