On Thu, 12 Jun 2014 05:54:47 -0700, Rustom Mody wrote:
> On Thursday, June 12, 2014 2:36:50 PM UTC+5:30, Steven D'Aprano wrote:
>> > The laws of physics tend to put
>> > boundaries that are ridiculously far from where we actually work - I
>> > think most roads have speed limits that run a fairly long way short
>> > of c.
>> "186,000 miles per second: not just a good idea, it's the law"
>> There's no *law of physics* that says cars can only travel at the
>> speeds they do. Compare how fast a typical racing car goes with the
>> typical 60kph speed limit in suburban Melbourne. Now compare how fast
>> the Hennessey Venom GT goes to that speed limit.
> Now you (or I) are getting completely confused.
> If you are saying that the Hennessey Venom (HV) is better than some
> standard vanilla Ford/Toyota (FT) based on the above, thats ok.
I'm not making any value judgements ("better" or "worse") about cars
based on their speed. I'm just pointing out that the speed limits on our
roads have very little to do with the speeds cars are capable of
reaching, and *nothing* to do with ultimate limits due to the laws of
Chris made the argument that *the laws of physics* put limits on what we
can attain, which is fair enough, but then made the poor example of speed
limits on roads falling short of the speed of light. Yes, speed limits on
roads fall considerably short of the speed of light, but not because of
laws of physics. The speed limit in my street is 50 kilometres per hour,
not because that limit is a law of physics, or because cars are incapable
of exceeding 50kph, but because the government where I live has decided
that 50kph is the maximum safe speed for a car to travel in my street,
rounded to the nearest multiple of 10kph.
In other words, Chris' example is a poor one to relate to the energy
efficiency of computing.
A more directly relevant example would have been the efficiency of heat
engines, where there is a fundamental physical limit of 100% efficiency.
Perhaps Chris didn't mention that one because our technology can build
heat engines with 60% efficiency, which is probably coming close to the
practical upper limit of attainable efficiency -- we might, by virtue of
clever engineering and exotic materials, reach 70% or 80% efficiency, but
probably not 99.9% efficiency. That's a good example.
Bringing it back to computing technology, the analogy is that our current
computing technology is like a heat engine with an efficiency of
0.000001%. Even an efficiency of 1% would be a marvelous improvement. In
this analogy, there's an ultimate limit of 100% imposed by physics
(Landauer's Law), and a practical limit of (let's say) 80%, but current
computing technology is so far from those limits that those limits might
as well not exist.
> In equations:
> maxspeed(HV) = 250 mph
> maxspeed(FT) = 150 mph
> so HV is better than FT.
"Better" is your word, not mine.
I don't actually care about fast cars, but if I did, and if I valued
speed above everything else (cost, safety, fuel efficiency, noise,
environmental impact, comfort, etc) then yes, I would say 250 mph is
"better" than 150 mph, because 250 mph is larger.
> But from your earlier statements you seem to be saying its better
> 250 mph is closer to 186,000 mps (= 670 million mph) than 150 mph
> Factually this is a correct statement.
And yet you're going to disagree with it, even though you agree it is
> Pragmatically this is as nonsensical as comparing a mile and a kilogram.
This makes no sense at all.
Your two statements about speeds are logically and mathematically
equivalent. You cannot have one without the other.
Take three numbers, speeds in this case, s1, s2 and c, with c a strict
upper-bound. We can take:
s1 < s2 < c
without loss of generality. So in this case, we say that s2 is greater
s2 > s1
Adding the constant c to both sides does not change the inequality:
c + s2 > c + s1
Subtracting s1 + s2 from each side:
c + s2 - (s1 + s2) > c + s1 - (s1 + s2)
c - s1 > c - s2
In other words, if 250mph is larger than 150mph (a fact, as you accept),
then it is equally a fact that 250mph is closer to the speed of light
than 150mph. You cannot possibly have one and not the other. So why do
you believe that the first form is acceptable, but the second form is
>> Speed limits for human-piloted ground-based transport ("cars") are more
>> based on social and biological factors than engineering ones.
>> Similarly, there are biological factors that force keyboards to be a
>> minimum size. We probably could build a keyboard where the keys were
>> 0.1mm square, but what would be the point? Who could use it? Those
>> social and biological factors don't apply to computing efficiency, so
>> it's only *engineering* factors that prevent us from being able to run
>> your server off a watch battery, not the laws of physics.
> As best as I can see you are confused about the difference between
> science and engineering.
> Saying one car is better engineered than another on direct comparison
> (150mph<250mph) is ok
> Saying one car is better than another because of relation to physics
> limits (c-150>c-250) is confusing science and engineering.
I do not understand what confusion you think you see here.
If we agree on the value judgement "greater top speeds are always
better", and the law of physics "c is the upper-limit to speeds", then
the following two statements are logically equivalent:
"Car HV is better than car FT because the HV has the greater top speed."
"Car HV is better than car FT because the HV's top speed is closer to c
than the FT's top speed."
These sorts of value judgments are independent of the *cause* of the
upper limit. Sticking to Chris' example of speed, if we agree that faster
travel is better than slower travel, then in the state of Victoria,
Australia, the ultimate upper-limit on (legal) speed is 110kph. If we
decide to value faster speeds, then the Hume Freeway with its 100kph
speed limit is better than my suburban back street with a 50kph speed
limit, even though the limit is a social restriction, not an engineering
limit or physics limit.
> Likewise saying AMD and Intel should have done more due diligence to
> their clients (and the planet) by considerging energy efficiency is
> right and I (strongly) agree.
> But compare their products' realized efficiency with theoretical limits
> like Landauers is a type-wrong statement
If I were arguing that there are no engineering limits prohibiting CPUs
reaching Landauer's limit, then you could criticise me for that, but I'm
not making that argument.
I'm saying that, whatever the practical engineering limits turn out to
be, we're unlikely to be close to them, and therefore there are very
likely to be many and massive efficiency gains to be made in computing.