On Fri, 28 Dec 2007 15:22:46 GMT, Robert Redelmeier
<redelm@[EMAIL PROTECTED]
> wrote:
>In comp.sys.ibm.pc.hardware.chips kony <spam@[EMAIL PROTECTED]
> wrote in part:
>>>No, they do not. Yes, the fast box benchmarks 3x. But I said
>>>"gets total work done that much faster". Does the user get up
>>>out of the chair in 1/3rd the time? No!
>
>> Yes, if we're goig to talk strict numbers then we would
>> have to say that since the faster system gets the job done
>> sooner, the user does get up sooner.
>
>Sure. Maybe 60 seconds on a one hour session.
.... or maybe 30 minutes in a one hour session. It does
little good to randomly pick numbers that suit your
argument, remember that not all uses are the same.
>1.7% for 3x the
>power. Not obviously a good return. But perhaps so in economic
>terms: Say +100W power draw at U$0.15/kWh is U$0.90/hr for the
>minute saved. That's cheap, and that's how energy is "wasted".
>
The processor is only a % of total system power. One can in
fact use a 2X faster processor with only 10W more power
consumed (at full load, less than 10W difference at idle if
as you suggest it would be mostly idle then the difference
is far less in practice).
So we might compare a system with 55W total consumption vs
60W total consumption when the latter is twice as fast.
Did you consider this? The slower system is always running
at a higher % load and further from it's idle state
consumption. The faster one is not running at peak power
consumption UNLESS it is fully utilized for a
compute-intensive job and if the job is that demanding there
really could be a difference in the time the system ran from
60 minutes down to 30 minutes.
>> There just isn't any thought beyond an overly simplistic "one
>> number is lower" concept here. It ignores all the reaslistic
>
>Chair time isn't realistic? I'm questioning whether the
>additional CPU power is useful, and arguing it may not
>be, leading to full life-cycle energy and resource savings.
Then the entire argument is invalid, because there is no
assumption a person who didn't find lesser performance
acceptible would have bought the new system using the higher
power figures. They'd be more likely using some Pentium 3
system that uses far less power, instead of buying two
addt'l systems.
I've been down this road already, my low powered system
didn't save enough power to matter.
IMO, the largest problem is the idea of splitting use
between two or more systems and having more hardware
manufactured, distributed, purchased, then eventually thrown
into a landfill. That does use power, REAL power, not
something we can ignore. Further we can't ignore that there
is a very good chance sometimes both systems would be
running, or even that the owner of the low powered system
would be less likely to turn it off, thinking it's not as
im****tant because it's *low power*.
In the end it just won't do much useful as a one-off idea.
Looking at our typical home energy bill it won't matter much
at all if an extra 10W were used. By ignoring this and
spending time thinking about trivial energy savings, we
allow larger energy consumers to be tem****arily ignored. It
wastes more power than it saves to even begin thinking about
this whole topic, because most people do not have some uber
gaming system with high power draw just sitting there
running 24/7. If they do, certainly I would suggest turning
it off or considering a secondary system but this is _not
the typical, reasonable comparitive scenario.
Have you tried using a system with a Via Eden processor and
Compact Flash or USB flash drive? I have, and even did a
lot of tweaking trying to improve performance like blocking
scripting and flash ads, things most people can't or don't
want to do. Let us know how well you feel it does for
general purpose use if you are using one, then compare it's
performance to what you could have for a mere 10-20W higher
power consumption with a different platform. I already
have, and the problem IMO is not conserving power, it's that
today the performance levels are not high enough _yet_. In
5 years maybe they will be, but during that period there are
a lot of other energy consumers that ought to be focused on
instead of the typical PC.
In fact, just switching to a smaller LCD monitor may save as
much or more power than the difference between the Via Eden
or C-7 platform and one twice as fast. Just putting on a
thicker ****rt and lowering the heater temp by 3 degrees may
help as much or more. Replacing lamps with some that don't
diffuse light as much might allow using fewer lamps. There
are tons of ways one could shave 10-20W off their energy
bill and still that is such a small difference it is a waste
of time in the larger picture.
You can't just cut a tiny bit of power usage then feel good
about it, that's just a token gesture that won't make any
real difference in the end once you consider all the
factors.
|
