Friday, May 22, 2009
Networks to get even more important
Previous versions of the, "End of Moore's Law" story have been superseded by technology breakthroughs. But this most recent one, by Saul Hansell, in the NYT Bits Blog, might have legs.
Eli Harari, the CEO of SanDisk says he anticipates about two more doublings, with doublings occurring about every year. “We can’t get below one [electron per memory cell]," he says.
Harari continues, “If I want 40 electrons, plus or minus two electrons, I can do that when the device is new. But seven years out, it will start to smear.”
He anticipates maybe two more doublings. So what it it's four? Or eight? "We can't get below one [electron]," is a fairly hard limit.
"When Manhattan ran out of space, they built skyscrapers," says Harari. What's a skyscraper in the world of bits? Hansell, entrained by the real estate analogy, imagines multi-layer chips. That's OK if you're only thinking chips.
In the wider digital economy, three things trade off with each other: storage, data compression and transmission. Data compression is the least general of these alternatives; it is costly, energy-intensive and/or inappropriately lossy. So, if Hansell's article is even close to right, I think we'll see new emphasis on the value of ubiquitous, inexpensive network connections.
Eli Harari, the CEO of SanDisk says he anticipates about two more doublings, with doublings occurring about every year. “We can’t get below one [electron per memory cell]," he says.
Harari continues, “If I want 40 electrons, plus or minus two electrons, I can do that when the device is new. But seven years out, it will start to smear.”
He anticipates maybe two more doublings. So what it it's four? Or eight? "We can't get below one [electron]," is a fairly hard limit.
"When Manhattan ran out of space, they built skyscrapers," says Harari. What's a skyscraper in the world of bits? Hansell, entrained by the real estate analogy, imagines multi-layer chips. That's OK if you're only thinking chips.
In the wider digital economy, three things trade off with each other: storage, data compression and transmission. Data compression is the least general of these alternatives; it is costly, energy-intensive and/or inappropriately lossy. So, if Hansell's article is even close to right, I think we'll see new emphasis on the value of ubiquitous, inexpensive network connections.
Technorati Tags: DisruptiveTechnology, Economics, NewYorkTimes, MooresLaw
Comments:
Post a Comment
