Slashdot Mirror
Why We Need a Second Moore's Law
Roland Piquepaille writes "In its April issue, Wired Magazine argues that we need a second Moore's law, this time about overall efficiencies of our computers and other electronic devices. The subtitle of the article summarizes it: "If we don't do something about increasing battery life, we're toast." Michael S. Malone, the author, says that the first Moore's law is endangered, not because the semiconductor industry cannot build new generation of chips, but because we will not be able to provide them with enough power. And he contends that the problem arises from the fact that we are using more and more wireless devices, which obviously are not connected to a plug. This overview contains selected excerpts of this eye-opening article."
Well if it's some day proven wrong, why just make another law that someday also might be inaccurate...
Increase battery life or decrease power consumption, or a combination of the two. The second is largely not consistent with the trend for increasing computing power of mobile devices, which only leaves the first. Not sure what a second Moore's law is supposed to do to help this necessary development.
a fair point, to be sure. although the article's author clearly feels that the basis for such a law exists in moore's paper, the fact of the matter is that i'm pretty certain engineers working on batteries are keenly aware of the need to develop more efficient batteries, capable of holding a charge longer, while ideally not becoming too hot to avoid "scorched lap" syndrome. further, i'm similarly certain that if such an observable principle actually existed that some reasonably perceptive engineer would have picked up on it and written his or her own dissertation along those lines.
we may need such a law, but at best, right now, it's just a bill.
ed
At the very leasy these gyms could push the generated power back into the power grid. They could even get paid for it by the power company.
Do I smell a new business model here?
really, springs, clocksprings to be exact. I brought this up in another discussion last week. I have two radios (baygen/freeplay)that make quite good use of windup to tighten a spring to run a microgenerator technology in lieu of batteries. I have another radio that has built in solar and a crank on the side that is a direct generator to on board rechargeable battery, plus it has another compartment that holds disposable batteries, or you can plug in a voltage adapter. It's an inexpensive radio, but it has 4 way power and works quite well. I understand now that grundig has an even higher quality radio with a similar crank to microgenerator scheme. This sort of technology makes use of extremely efficient energy conversion and energy storage, ie, biochemical from the human body, that beats heck out of any battery out there. How about at least starting with a PDA to see if the windup style concepts have merit and can be adapted up the useage scale then? I see a lot of these PDAs use AA or AAA batteries, the same as these small radios, seems a natural to me. Even just a power adapter that is the spring, crank and battery bank, and that plugs into existing PDAs if they have a DC jack in. something along those lines. It's just not that hard to run a tiny crank for 30 to 60 seconds.
You know what I want someone to invent for me? A stationary bicycle that's nothing but the pedals and some sort of mechanism for adjusting the resistance, that can be placed under a standard sized desk. I'm the Queen of multi-tasking, and it would be great if I could somehow give my feet something to do, especially something physically constructive, while my hands and brain were working on other things.
I'd make it myself but I'm too busy drawing and don't really know anything about making stuff like that.
My Webcomic: Asylum on 5th Street
Moore's first law is a two-edged sword - more transistors for the same price is great for computers, but it's hell on batteries: As the processor power doubles, the power consumption also rises.
Correct me if I'm wrong, but: Malone, the author, is exaggerating by implying that the size of a transistor is remainaining constant while the number of transistors doubles. As I understand it, the smaller the processor, the less power it requires. Is this right?
Sure the chip industry needs to work on energy usage (perhaps through either fuel-cell batteries for lap-tops). Also, Malone is merely following the wagon with Intel's recent processor naming change. They've already figured out, that cycles are losing their prior applicability.
I don't think there is a problem with power usage here. Cellphones have the power to be full fledged telephones, electronic organizers and even office computers (mail, word processing, web browsing, etc.) all in one. Yet, one cellphone battery charge lasts a lot longer than a few years ago. And I think noone would argue cellphones aren't wireless.
If cellphones can do it in such small form factors, why wouldn't larger devices like notebooks be able to do the same? I know that most pc-compatible notebooks are engineered for speed, not battery life, but look at Apple's, for example. They live for more than 5 hours (and they really do) on one charge, which I think is quite respectable.
Please correct me if I got my facts wrong.
Yes, wireless devices are becoming more prevelant, but they will never fully replace wired fixed workstations. While we can always work towards wireless devices that use less power, better batteries, and better wireless connectivity, it will not match the speed and power of a desktop. The idea that battery life is going to limit the semiconductor industry is foolish.
DeviantArt Page
NSFWThis is such low quality news, it's depressing. It's the kind of news that's formulaic: "what can I write about, oh, I know, let's take an age old thing, like e=mc^2, or Moore's law because computer geeks prefer that one, and then use it to spin story on an age old problem ... energy conservation".
yay.
Brilliant really.
Moore's law, and the reason for its brilliance is that the guy was so right... it's that for 30 years, people have been saying it's gonna fall soon, and yet the law still holds invariant, like some forteress somewhere.
That's the brilliance of a law: that it's invariant. You don't make brilliance by adding "oh and this"-exceptions.
IMHO, Moore's law will still not be broken, or even hurt by these energy consumption problems for at least 10 more years (no pun intended). Just check out the new tech they came up with for chip size in water immersion (news from yesterday on /.). There will be revolutions made in power as well... don't you doubt it, no matter what some Wired news reporter would want you to think.
Amen. I learned so much from working on small, underpowered systems. Having to think about where your CPU cycles and disk space are going forces you to pick up much more of a feel for how things really work. (e.g. I'm running out of space in /var - let's see what all those files are *really* for. What packages can I lose? What can I turn off or tweak? Is there a better way of doing this?)
Eventually you acquire a low-level "feel" for what the machine's doing, and that's how you're able to fix problems later along the line. I know it doesn't matter in the long run, but I *hate* seeing inefficient use of resources. Even things like scripts using awk when they could use cut piss me off. It's lazy. Understand your craft.
Over the last ten years my computers have got faster and faster. (Also hotter and hotter and louder and louder.) Have I got more and more productive? No. I can't write a letter any more quickly with Word on a 2+GHz XP machine than I could with Wordworth on my 30MHz Amiga in 1992.
I haven't replaced my personal machine in about five years, and I've got no plans to do so any time soon. Even 20GHz won't make me write quicker, read quicker or think quicker. I'm more interested in machines running cooler, quieter and cheaper.
You are so right! In fact, I'd say that Moore's Law was far from a new idea. It applies to just about everything and anything. Remembering back to Econ 101 (albeit cloudily so) many people back in the 18th and 19th centuries were freaked out that soon there wouldn't be enough land for crops to feed the increasing populations. And they were right...based on current conditions. What they never took into consideration is the advancement in technology which provides higher and higher yields on less land. He's making arguments almost as if battery technology will stand still while processor power keeps increasing.
And let's not forget that necessity is the mother of all invention and of course the profit motive of every corporation out there. If there's money to be made, things will happen. It's the immutable law of capitalism, which isn't that much different from Moore's Law.
Everybody dies frustrated and sad and that is beautiful