Slashdot Mirror
Whither Moore's Law; Introducing Koomey's Law
Joining the ranks of accepted submitters, Beorytis writes "MIT Technology review reports on a recent paper by Stanford professor Dr. Jon Koomey, which claims to show that the energy efficiency of computing doubles every 1.5 years. Note that efficiency is considered in terms of a fixed computing load, a point soon to be lost on the mainstream press. Also interesting is a graph in a related blog post that really highlights the meaning of the 'fixed computing load' assumption by plotting computations per kWh vs. time. An early hobbyist computer, the Altair 8800 sits right near the Cray-1 supercomputer of the same era."
My favorite example of computing (in)efficiency is the USAF's SAGE bomber tracking computers introduced in the 1950s. These vacuum tube machines had CPU horsepower probably in the same ballpark as an 80286, but could draw more than 2 megawatts of power each. They didn't decommission the last one until the 1980s.
Yes, there is if you "erase" intermediate results -- look up 'von Neumann-Landauer limit', kT*ln(2) energy must be dissipated for non-reversible computation.
Reversible computation can theoretically approach zero energy dissipation.
Wikipedia is your friend! :)
Paul B.
Since the energy is only required when information is erased, reversible computing can get around this requirement. Aside from basic physics-level problems with building these logic gates, the problem with reversible computing is that it effectively requires keeping each intermediate result. Still, once we get down to anywhere close to the kT ln 2 physical constraint, reversible logic is going to look very attractive.
"Evil company X is threatening to restrict our rights! Let's all get together to stop--OOOH! SHINEY!!!" -- AC
I do not think that you get net energy savings (by using the same basic technology, e.g., CMOS at room temeprature or "cold"), if you take into account the fact that cooling things down also costs energy! For example, liquid helium refrigeration costs about 1 kW of wall outlet power to compensate for 1 W dissipated at 4.2 K.
Changing your basic technology to, e.g., some version of superconductor-based logic can help (a lot!), current state of the art (in my very biased opinion, since I am cheering for those guys, and have been involved in related research for years) is here: http://spectrum.ieee.org/semiconductors/design/superconductor-logic-goes-lowpower ...
Paul B.
There's an even more obvious difference.
The Cray-1 is sitting half a division above the line. As that's a logarithmic abscissa, that Cray is putting out about 3X as many calculations per KWh as the on-the-line entrants are.
The Altair-8800 is sitting right on the line, being non-impressive to its contemporaries, while the Cray is blasting them with its laser vision and eating nothing but salads.
Yes, reversible computation can theoretically approach zero energy dissipation, but if you use no energy, the computation is just as likely to run forwards as backwards. You still need to consume energy to get the computation to make progress in one direction or the other. Richard Feynman has a good description of this idea in his Lectures on Computation.
What a fool believes, he sees, no wise man has the power to reason away.
What amazes me is the computation done in biological systems.
When I consider the amount of correlation and replication done by RNA/DNA systems, I am left in the dust, wondering just what happened.
I'm not sure I would classify a polymerization as a "computation." Even then the RNA transcription rate is on the order of ~50 nucleotides per second or so, which isn't all that stunning. The only thing that's really impressive is how interdependent the chemical reactions are, and how sensitive the whole system is.
Don't be fooled by the DNA :: Computer Code analogy - it is very, very wrong.
=Smidge=
Most likely what just happened is you got laid.
C|N>K
It's the inverse of Moore's law so yeah, duh....
If your compute power doubles in the same size die every 1.5 years, then if you halve the die size keeping the compute power the same you actually cut the power in half. This is a very well known phenomenon and Koomey is doing what he has been for a while, making headlines with little substance and lots of flair.
That Microsoft and Intel paid for this research calls into question what it was they were actually paying for.
Who are you? The new #2 Who is #1? You are #617565. I am not a number, I am a free man! Muhahaha.
What about the energy used creating efficiency?
Are we experiencing an increase in efficiency?
OR
Are we expending every increasing amounts of energy creating the appearance of efficiency?
In the distance you hear an ominous moo.
The thing is that even if we could do the whole calculation using reversible computing, then what? If we start over on a new and completely different calculation we can't use any of the previous intermediaries and if we clear them - either before or during the next calculation - then we've just spent as much energy as doing it the non-reversible way. Reusing past calculations or lookup tables that are really cached results is something we do in many algorithms today, so each calculation is likely to be necessary and then I don't see how reversible computing is going to do anything but fill the computer with useless intermediaries. We just delay the energy use until we somehow dispose of or reset them.
Live today, because you never know what tomorrow brings
55,000 tubes vs. 134,000 transistors
Had 256 KB + 16 KB RAM vs. the 512-640 KB common in the 286
75,000 instructions per second vs. 1.2 million (@6 MHz)
SAGE used 52 of them, half online at a time, geographically dispersed, all working on tracking aircraft. But they did communicate with each other, so you might consider this a 1,950,000 instructions per second cluster, beating the first 286s that came out around the time SAGE was stood down.