is power consumption proportional to workload?

Does power consumption of the 8800gt vary significantly with workload, or is
it fixed?

I'm about to build a new PC. I'm not a gamer at all, but would like to be
able to run Flight Simulator X (FSX). That will be on rare occasions, but
I'd like it to run well when I do. I've read that an 8600gt consumes about
43 watts, but the 8800gt consumes 138 watts -- a huge difference. (My
reference is
en.wikipedia.org/wiki/GeForce_8_Series ) 8800gt prices are coming down and
I'd be willing to pay the difference over an 8600gt for a good FSX
experience, but am not willing to pay for the power (or suffer the heat
thereof) of the extra 100 watts continually.

However, if these power figures are under maximum load, which would occur
under gaming, and if the power using normal office applications would be
greatly less than those figures, then I'd go with the 8800gt. Otherwise,
I'd go with the 8600gt.

Can anyone answer this definitively, please?

'BobAtVandy' wrote, in part:
Does power consumption of the 8800gt vary significantly with workload, or
is it fixed?
_____

Yes.

That is the definitive answer. But you probably want to ask the question in
a different way.

Phil Weldon


"BobAtVandy" wrote in message
news
Does power consumption of the 8800gt vary significantly with workload, or
is it fixed?

I'm about to build a new PC. I'm not a gamer at all, but would like to be
able to run Flight Simulator X (FSX). That will be on rare occasions, but
I'd like it to run well when I do. I've read that an 8600gt consumes
about 43 watts, but the 8800gt consumes 138 watts -- a huge difference.
(My reference is
en.wikipedia.org/wiki/GeForce_8_Series ) 8800gt prices are coming down
and I'd be willing to pay the difference over an 8600gt for a good FSX
experience, but am not willing to pay for the power (or suffer the heat
thereof) of the extra 100 watts continually.

However, if these power figures are under maximum load, which would occur
under gaming, and if the power using normal office applications would be
greatly less than those figures, then I'd go with the 8800gt. Otherwise,
I'd go with the 8600gt.

Can anyone answer this definitively, please?

BobAtVandy wrote:
Does power consumption of the 8800gt vary significantly with workload, or is
it fixed?

I'm about to build a new PC. I'm not a gamer at all, but would like to be
able to run Flight Simulator X (FSX). That will be on rare occasions, but
I'd like it to run well when I do. I've read that an 8600gt consumes about
43 watts, but the 8800gt consumes 138 watts -- a huge difference. (My
reference is
en.wikipedia.org/wiki/GeForce_8_Series ) 8800gt prices are coming down and
I'd be willing to pay the difference over an 8600gt for a good FSX
experience, but am not willing to pay for the power (or suffer the heat
thereof) of the extra 100 watts continually.

However, if these power figures are under maximum load, which would occur
under gaming, and if the power using normal office applications would be
greatly less than those figures, then I'd go with the 8800gt. Otherwise,
I'd go with the 8600gt.

Can anyone answer this definitively, please?




If you are that concerned about the possibility of using an extra 100W a
few hours a day you need to invest in a cheap kill-a-watt device
(Google) it can measure and log effortlessly.

You can then find out your actual electrical usage over time...say a
week or so and using the supplied information and your KWH/$.$$ rate
determine what steps you need to take, such as power saving modes, to
meet your goals.

Idle wattage does differ among cards and PC configurations, there are
numerous PC Power Calculators (Google) which can approximate your idle
wattage based on your input.

As an example, currently reported by a Kill-A-Watt:

idle: 140W
3DMark2006: 235W
8800GTS(G92)512M @ 700MHz
OPT185 @ 2.8GHz
2G DDR
1 HDD
3 Fans
2 PCI cards + PC Powered Breakout box
12 USB devices with 2 being AC powered.

The maximum charge so far in my zone is 0.16/KWH (including all KWH
sensitive fees).

Taking the 95W an hour difference between idle and ready and gaming and
doing the 16 cents a KW hour math gets you a grand total of 2 cents an
hour extra to game than to idle.

If you gamed two hours a day every day. At the end of a year you would
pay $11 more in electricity than if you did not. We can ignore the
speakers and network equipment's current draw unless you only use them
for gaming, but I bet you do not.

Lastly, the 8600GT sucks don't buy it unless you want people to point at
you mockingly.

BobAtVandy wrote:
Does power consumption of the 8800gt vary significantly with workload, or is
it fixed?

I'm about to build a new PC. I'm not a gamer at all, but would like to be
able to run Flight Simulator X (FSX). That will be on rare occasions, but
I'd like it to run well when I do. I've read that an 8600gt consumes about
43 watts, but the 8800gt consumes 138 watts -- a huge difference. (My
reference is
en.wikipedia.org/wiki/GeForce_8_Series ) 8800gt prices are coming down and
I'd be willing to pay the difference over an 8600gt for a good FSX
experience, but am not willing to pay for the power (or suffer the heat
thereof) of the extra 100 watts continually.

However, if these power figures are under maximum load, which would occur
under gaming, and if the power using normal office applications would be
greatly less than those figures, then I'd go with the 8800gt. Otherwise,
I'd go with the 8600gt.

Can anyone answer this definitively, please?



This site has measured power consumption. They install current shunts in
the PCI Express slot, and also measure the Aux power connector, and do
it for various test conditions. Near the bottom of the page, is a comparison
chart, comparing various cards of a similar vintage.

http://www.xbitlabs.com/articles/vid...-8800gt_5.html

In this graphic, "idle", the table on the left, is what the 8800gt will
burn sitting in the Windows desktop.

http://www.xbitlabs.com/images/video...power_full.gif

The "3D peak" would be in a gaming situation. I'm not really sure what
their "2D peak" would correspond to, in terms of real world usage.
Perhaps scrolling a document, or doing something that involves a lot
of bitblt or something.

Note that these numbers are measured for *one* card, and thus if a
statistician was involved, would be invalid for the purposes of
establishing how all cards behave. When Nvidia states a value for
power, it might be some kind of engineering max, and hard to
relate to an Xbitlabs measured value.

Paul

"Phil Weldon" wrote in message
...
'BobAtVandy' wrote, in part:
Does power consumption of the 8800gt vary significantly with workload, or
is it fixed?
_____

Yes.

That is the definitive answer. But you probably want to ask the question
in a different way.

Phil Weldon



I disagree. I would say the definitive answer is No.

Ed