Re: Low power software + high power hardware = high power dissipation?
- From: Paul <nospam@xxxxxxxxxx>
- Date: Wed, 13 Feb 2008 22:35:21 -0500
blackhead wrote:
Suppose you have Word2000 which happily ran on a PentiumII/AMD socket
7 system and you now put it on a high powered system with exactly the
same operating system and software installed. The power dissipation is
higher and all because the software is being run faster when it
doesn't need to be. Whereas in the past it might check for a user
response from the mouse or keyboard y times a second, now it's x*y
where x is how much much faster the system is. It's ridiculous, isn't
it?
Is this problem being confronted now, so that software has a speed
rating for all its modues that define how fast they are needed to run?
Module responses requiring user input via the mouse or keyboard might
have a rating of 20 cycles per second, whereas the module that spools
a file for a printer could be done as fast as possible since it is
done only once and doesn't recycle. This way, as systems become
faster, they become more efficient by going into sleep or idle mode
when a module doesn't require servicing yet.
Interested in your views
If I were to do a fresh install of the OS, and had no software running,
my processor (3GHz Northwood) would use 13.2 watts at idle. That is
measured with a clamp-on ammeter, and it sees 1.1 amps being drawn
from the 12V rail.
There are several features, which make low power operation like that possible.
1) Programs are interrupt driven, and don't sit in a polling loop.
Only dumb programmers use polling loops. Modern programs "block on I/O",
and the program doesn't do anything, until a significant event, like
a key press or mouse click occurs.
2) The operating system has an "Idle" task. The scheduler runs the Idle task,
when no other process is ready to run. The Idle task can be coded with a
halt (HLT) instruction in it, which will actually cause the processor to
go to sleep. What wakes the processor up, is any hardware interrupt. An
example of a regularly occurring interrupt, is the "Clock Tick Interrupt".
In fact, the Clock Tick is what triggers the Scheduler to run, and then the
Scheduler can check again, to see what process is ready to run. If
nothing is ready to run, then the Idle task runs for another time slice.
3) Features like Cool N' Quiet or Intel Speedstep (EIST), allow Vcore and
Fcore to be dropped, when the OS detects that the load on the system is
low. By using those features, it is possible to further reduce the
operating power.
Can something screw that up ? Yes.
I installed some Antivirus software, and it appears to replace the system
Idle loop with its own. The antivirus software doesn't have a HLT instruction
coded in the Idle loop. This causes the processor to run hotter than it
would otherwise do. (I just checked a few minutes ago, and the processor is
using 54W or 12V @ 4.6amps, and that is due to the antivirus software.)
And there is nothing to see in Task Manager, when HLT is not being used.
Only my power measurement device, can tell the difference.
If I download a copy of RMClock, RMClock can install its own Idle loop, and
it does use a HLT instruction. If I choose to run RMClock right now, my
CPU power drawn drops back to 13.2 watts. So, in this example, there
were three Idle loops involved. The OS has a good one (13.2W). The
antivirus software has a bad one (54W). RMClock has a good one (13.2W).
Naturally, the antivirus software squawks a bit, when I install RMClock :-)
If you want to build a low power computer, there are processors
that would be good for the job. For example, this Conroe-L is 35W
max, and draws a lot less when idle. This would be a good match for your
Pentium II generation processors. It is possible to build a modern
system, that uses less power under all conditions, than your
Pentium II. (Putting a high end video card in the system,
doesn't help. Video can be pretty wasteful. Integrated graphics
can be better, but some chipsets run cooler than others. )
http://www.newegg.com/Product/Product.aspx?Item=N82E16819116039
HTH,
Paul
.
- Follow-Ups:
- References:
- Prev by Date: Re: how to delete virus in system volume info?
- Next by Date: Intel price cuts
- Previous by thread: Re: Low power software + high power hardware = high power dissipation?
- Next by thread: Re: Low power software + high power hardware = high power dissipation?
- Index(es):
Relevant Pages
|