A Conversation with Alan Kay (2004)

queue.acm.org

12 points·mercer··4 comments

> Just as an aside, to give you an interesting benchmark—on roughly the same system, roughly optimized the same way, a benchmark from 1979 at Xerox PARC runs only 50 times faster today. Moore’s law has given us somewhere between 40,000 and 60,000 times improvement in that time. So there’s approximately a factor of 1,000 in efficiency that has been lost by bad CPU architectures.

Anyone know what kind of benchmark he's talking about? Because if that's true, holy shit.

(considering of course, that this article is from 11 years ago)

Well ... Moore's Law is not about speed, but the number of low cost transistors you can economically put on a single die. How that translates into "speed" is complicated, e.g. when you're first able to put all of a CPU on a single die you get a big speedup. After this was written raw clock speeds hit a wall, all that's left is microarchitecture improvements, including making things wider, making dies wider by putting more cores on them, and adding lots of cache.

Hmmm, and a whole lot of architecture work has by and large ended up on the garbage heap of history, or is no longer trying to compete in speed. Plus ARM has always prioritized power over speed.

Let me guess...object oriented programming was more about message passing than anything else.

I'm just kidding, I love reading Alan Kay. He's awesome.

If you want to read a terrific interview in which he touches on WWII's effect on him, how Socrates could get to heaven, how programming has turned to pop culture, and of course message passing, see his interview in Dr. Dobb's[1]

[1] http://www.drdobbs.com/240003442

Reply on news.ycombinator.com