Gordon Moore predicts end to Moore's law in 10 years
By Thomas Ricker 
posted Sep 19th 2007 3:16AM
posted Sep 19th 2007 3:16AM
After more than 40 years of empirical truth, Moores law -- the maxim which declares a doubling of transistors on a computer chip roughly every two years -- is under attack... by Gordon Moore himself. Ok, he's just the messenger in this case; it's the laws of physics that will render Moore's Law obsolete in "another decade, a decade and a half" according to the co-founder of Intel. Fact is, space on a chip is finite so eventually, (this isn't the first time he's predicted the end) he'll be right. Still, in perhaps a divination of future processing leaps, Moore noted that "the interface between computers and biology now is a very interesting area." Yes Gordon, we're all waiting for a quantum computer powered by a teaspoon of bacterial goo. Well, that, jetpacks and flying cars.
What - you never saw this?
-Green, leafy spinach may soon power more than Popeye's biceps -http://web.mit.edu/newsoffice/2004/spinach-0915.html
Keep'n it real!
He & I know more on this one than the gadget freaks, whoops, fans!
J.
Speaking of incredibly practical articles from MIT, why don't we power the spinach computer with wireless electricity?
http://web.mit.edu/newsoffice/2007/wireless-0607.html
Soylent Green: Entire computers that are edible if the going gets tough.
Clarke's First Law: “When a distinguished but elderly scientist says something is possible, he is almost certainly right. But when he says something is impossible, he is very probably wrong.” http://www.lsi.usp.br/~rbianchi/clarke/ACC.Laws.html
Mock ye not. . Keep your comments about green goo powering quantum computers. It will ground you in the future when it happens. Remember, science fiction never imagined the pc . . green goo we can.
If we switch to green goo, then we make the move away from transistors on a chip. Which proves Moore right. A major shift in technology doesn't contradict this prediction.
Who needs green goo? Our bodies, and those of other organisms around us, are likely to be the future of computing: organically powered, portable, always on (in a processing sense).
ummm, i don't like Green. Can I get the pink goo computer instead. with glitter. I like glitter
So this would be Moore's second law.. neat.
If Moore's law has a limit, what will our future robot overlords use?
Googlephones
I always thought Moore's law was a part of a greater law. sure we will run out of space on silicon aparently in like 2015 or so but there are other possibilities like 3D computing and biological computing that will take over by then.
anyone interested in this should check out the futurist Ray Kurzwiel.
Silicon chips are already stacked vertically, the circuits criss-cross, the biggest problem with it is signal bleed between the layers. Which mandates shielding, which adds to the size without increasing performance.
biological computing ... think I'll just slip my dead aunt's gray matter in the back of my PC and fire it up!
I'm not so sure... chips are starting to go more 3-dimensional these days, which could easily lead to quite a few more years of increased capacity. Once we hit the practical limit for 2D, start stacking things upwards towards another practical limit. By that time we should be working with optical processors, and the race will continue with different (and smaller) physical limits.
I dont see where this is a problem, most software isn't written well enough to take advantage of the processing power that is avail now with dual/tri/quad core chips. Let alone the 7/8 core cell. Which sony was blaming software programming for not using its full potential.
To relate to a car, why build a super horse power motor and the put a govenour on it?
sorry I know i didn't spell that right. Can someone program a spell checker into browsers?
Try Firefox... spell checking is built into the browser... just right click and choose the correct word you were trying to spell. Incidentally, it is spelled governor.
I love the humor in the engadget writeups :-)
Sure there is a possibility of those quantum computers and bio computers being developed and becoming mainstream but it has so many iffs.
How about parallel processing? Getting into hundreds of cores and writing programs that support the multicores?
After all when we are talking about supercomputers available there now in a nutshell all they are a giant multicore.
Why would you want to put a super computer in a nut shell?
Hey that would be nice to have it that small :)
"In a nutshell," means "in a few words," or "very briefly explained." Nutshells, being the "hard exterior within which the kernel of a nut is enclosed" (to quote the Oxford English Dictionary), don't get very big since nuts themselves are generally fairly small.
Well, what's wrong with hitting the end of how many transistors you can pack on a piece of silicon? There are still other options. First off, you can always add more processors and make the piece of silicon bigger, replace mechanical drives/CD and DVD ROMS with silicon based drives and we'll have much faster systems. Then you have optimization of the OS and applications.
IMO all of the computers/laptops we have been using have all been prototypes of what will eventually be the ultimate computer/laptop and then we can just focus on using the darn things and stop arguing over which one is better.
I personally think that eventually we'll have only one microprocessor company that makes all of the various types of microprocessors we'll ever need. Eventually pure capitalism is going to prove that it doesn't work, just like pure socialism doesn't work.
Maybe between now and 10 years from now the Mac OS will become perfect and we won't have to deal with Windows. Maybe Mac OS X is ultimately going to be perfected where it won't need updates as it will have all of the features we'll need and it will be fully optimized and it won't crash and in that time frame all of the applications will have all of the features we'll ever need and we won't have to keep on upgrading.
Then we'll have time to just focus on actually using the computers and applications and having fun without having to deal with fixing them.
The only problem with that particular train of thought, in its current iteration, is that it completely misses the most vital piece of the entire argument: the affect of the end user.
As a net admin in education, I have personally witness end users, who honestly believed they were doing 'the correct thing' blow their systems sideways in some SPECTACULARLY amazing ways... ways I cannot even begin to explain here. So until we solve the 'end user dilemma', as it were, no advance in the computing industry will ever prevent us from not having to fix software or hardware. All we can really hope to do is make it faster and 'candy coated' enough that they can't really hurt themselves as badly when they start tinkering with it under the hood. This very thing has a lot to do with how OS X and Vista are put together, especially Vista. Microsoft has become weary of being the public whipping boy every time their end users blow up their home or work machines doing things they expressly should not do, things that the rest of us would typically consider common sense.
Apple is on its way of perfecting the OS far faster than Microsoft and eventually there won't be any end loser issues. I think natural selection might either rid ourselves of people that don't know how to use computers, or they'll get proper training or the software/OS will become easier to use and less problematic. Then we won't need as many support personnel.
Here's the future
http://www.youtube.com/watch?v=ac0E6deG4AU
I saw a demo similar to this a while ago, but this is a great video. Personally, I think this technology is coming to laptops and desktops with better apps maybe in the next year or two. The number of apps that take advantage of this will be small, but growing. From what I have seen in the industry, there is probably about 1% of the tens of thousands of applications that are actually worthwhile. There are a lot of apps that should be built right into the OS and that has been gradually happening over the last 20+ years. I think that trend will be increasing at a faster rate. At least Apple bundles usable apps when you buy a Mac.
This will force programming in general not to rely on brute force of future computing power and start optimization of software.
...it will sure trump Moore's law that was applicable to programming where the amount of code in programs was doubling at the rate equivalent to the silicon formula.
This could be a good thing. Instead of getting faster and faster they'll be getting cheaper and cheaper. Hopefully.
They are getting cheaper and cheaper. I believe the law actually has a monetary value involved.. We are not actually doubling our computing power we are doubling the computing power at a certain price point which is ~ $1000.
i realize that. But when the newer ones come out the others tend to get discontinued so the price points don't actually lower.
Gordon Moore interview with Moira Gunn at IDF 2007 is available online. Do a Google search on "Gordon Moore Video".