Idle powerusage 6200 vs. 8400

Lasse Jensen wrote:
> Some time ago i saw an article stating that Geforce 8400 used less power
> when idle than 6200. I can't seem to find the source anymore. So i brought
> a no-name 8400 GS a couple of weeks ago. The computer i use the card is
> always turned on, but rarely used for any gaming. Anyway, the card runs
> really hot, even when it's not doing anything besides showing a 2D desktop.
> Is it a possible driver issue? I use the latest binary Linux driver from
> Nvidia. Does anyone know where i can compare how much power the two card in
> question should use?
>


Kill-A-Watt

http://www.amazon.com/P3-International-Kill-Electricity-Monitor/dp/B00009MDBU

or

http://www.thinkgeek.com/homeoffice/gear/7657/

Accurate, cheap, easy to easy, informative, does what you want and more.
 >> Stay informed about: Idle powerusage 6200 vs. 8400 

Phil Weldon wrote:

> 'Lasse Jensen' wrote:
> | Some time ago i saw an article stating that Geforce 8400 used less power
> | when idle than 6200. I can't seem to find the source anymore. So i
> | brought a no-name 8400 GS a couple of weeks ago. The computer i use the
> | card is always turned on, but rarely used for any gaming. Anyway, the
> | card runs really hot, even when it's not doing anything besides showing
> | a 2D
> desktop.
> | Is it a possible driver issue? I use the latest binary Linux driver from
> | Nvidia. Does anyone know where i can compare how much power the two card
> in
> | question should use?
>
> Numbers, please.
>
> The temperature of the GPU is dependent on heat produced AND the cooling
> solution AND the temperature of the air entering the heatsink (if air
> cooled.)
>
> Also, if a fan is used, the idle temperature will also be dependent on the
> fan speed (which, on your particular card, may be set at zero when the
> video card is not being heavily used and the GPU temperature is below 70
> C.

Both cards are fanless.

> Also, as for a 'really hot' GPU being a problem, it all depends on what
> you mean by 'really hot'.

Really hot meaning this device is gonna cost a lot of money in electricity.
With the 6200 i was able to touch the heatsink for a longer period of time.
With the 8400 i'm not. If it weren't for the occasional quake 4 multiplayer
games i would go back to a Matrix Millennium PCI card

> GeForce 8400 GS:
> 38 Watts power consumption (probably
> power
> consumption during heavy 3D acceleration, judging by the numbers given for
> the GeForce 8800 GTX)
> 80 nm process GPU
> 210,000,000 transistors
> 450 MHz GPU clock
> according to
> http://en.wikipedia.org/wiki/GeForce_8_Series .
>
> GeForce 6200? WHICH GeForce 6200?

NV44 with 128 mb memory on board.

> GeForce 6200:
>
> 110 nm process GPU
> 77,000,000 transistors
> 350 GPU clock
> according to
> http://en.wikipedia.org/wiki/Geforce_6200 .

Smaller manufacturing process, 3 times more transistors and a slightly
higher clock frequency.

> The report at
> http://www.xbitlabs.com/articles/video/display/ati-vs-nv-power.html#sect0
> is an example of how video card power consumption SHOULD be measured, as
> opposed to the many reports that just show the total system power
> consumption ( numbers which are of limited use since exercising a 3D
> accelerated video card is can also be expected to exercise the rest of the
> system.

None of the cards have external power connectors. But it would be the best
way of measuring it. I don't have the right gear to produce dual layer
PCB's in the quality needed, so building a riser card is also out of the
question. Although it would be the best way of measuring it.

--
Lasse Jensen [fafler at g mail dot com]
Linux, the choice of a GNU generation.
 >> Stay informed about: Idle powerusage 6200 vs. 8400 

'Lasse Jensen' wrote, in part:
| Really hot meaning this device is gonna cost a lot of money in
electricity.
| With the 6200 i was able to touch the heatsink for a longer period of
time.
| With the 8400 i'm not. If it weren't for the occasional quake 4
multiplayer
| games i would go back to a Matrix Millennium PCI card
_____

Well, you could use ATITool to underclock the GeForce 8400 GS or nVidia
Control Panel to use profiles based on GPU demand (assuming that Linux also
has useful versions of these utilities.)

You might want to look into the nVida controls for mobile GeForce 8X00 video
adapters. My notebook has a GeForce 8400M GS and nVidia drivers that have
the equivalent of Intel 'Speed-Step'; three different sets of clock rates
for 2D, light 3D, and heavy 3D requirements.

But numbers are important. A C-7 candelabra base 'traditional' Christmas
tree lamp consumes only 7 1/2 Watts, but you wouldn't want to hold one for
very long; showing the importance of the cooling solution. Since your video
card has a passive cooler, then 'really hot' may not mean much in the way of
power consumption. Especially if your case has minimal air flow.

You might find a non-contact thermometer useful for getting numbers. Harbor
Freight has a non-contact infrared digital thermometer for ~ $3 US; and
throws in hold peak and stopwatch functions.

As for electrical power expense, the GeForce 8400 GS @ 38 Watts peak
consumption would cost, in Atlanta, Georgia, only ~ $11 US for 38 Watts X 10
hours X 365 days X $0.08/ KWH (and that in a location where less than 30% of
electrical power is produced with other than fossil fuels.) Any monitor you
use will consume much more electrical power.

Phil Weldon



"Lasse Jensen" <fafler RemoveThis @gmail.com> wrote in message
news:47a585fb$0$89176$157c6196@dreader1.cybercity.dk...
| Phil Weldon wrote:
|
| > 'Lasse Jensen' wrote:
| > | Some time ago i saw an article stating that Geforce 8400 used less
power
| > | when idle than 6200. I can't seem to find the source anymore. So i
| > | brought a no-name 8400 GS a couple of weeks ago. The computer i use
the
| > | card is always turned on, but rarely used for any gaming. Anyway, the
| > | card runs really hot, even when it's not doing anything besides
showing
| > | a 2D
| > desktop.
| > | Is it a possible driver issue? I use the latest binary Linux driver
from
| > | Nvidia. Does anyone know where i can compare how much power the two
card
| > in
| > | question should use?
| >
| > Numbers, please.
| >
| > The temperature of the GPU is dependent on heat produced AND the cooling
| > solution AND the temperature of the air entering the heatsink (if air
| > cooled.)
| >
| > Also, if a fan is used, the idle temperature will also be dependent on
the
| > fan speed (which, on your particular card, may be set at zero when the
| > video card is not being heavily used and the GPU temperature is below 70
| > C.
|
| Both cards are fanless.
|
| > Also, as for a 'really hot' GPU being a problem, it all depends on what
| > you mean by 'really hot'.
|
| Really hot meaning this device is gonna cost a lot of money in
electricity.
| With the 6200 i was able to touch the heatsink for a longer period of
time.
| With the 8400 i'm not. If it weren't for the occasional quake 4
multiplayer
| games i would go back to a Matrix Millennium PCI card
|
| > GeForce 8400 GS:
| > 38 Watts power consumption (probably
| > power
| > consumption during heavy 3D acceleration, judging by the numbers given
for
| > the GeForce 8800 GTX)
| > 80 nm process GPU
| > 210,000,000 transistors
| > 450 MHz GPU clock
| > according to
| > http://en.wikipedia.org/wiki/GeForce_8_Series .
| >
| > GeForce 6200? WHICH GeForce 6200?
|
| NV44 with 128 mb memory on board.
|
| > GeForce 6200:
| >
| > 110 nm process GPU
| > 77,000,000 transistors
| > 350 GPU clock
| > according to
| > http://en.wikipedia.org/wiki/Geforce_6200 .
|
| Smaller manufacturing process, 3 times more transistors and a slightly
| higher clock frequency.
|
| > The report at
| >
http://www.xbitlabs.com/articles/video/display/ati-vs-nv-power.html#sect0
| > is an example of how video card power consumption SHOULD be measured, as
| > opposed to the many reports that just show the total system power
| > consumption ( numbers which are of limited use since exercising a 3D
| > accelerated video card is can also be expected to exercise the rest of
the
| > system.
|
| None of the cards have external power connectors. But it would be the best
| way of measuring it. I don't have the right gear to produce dual layer
| PCB's in the quality needed, so building a riser card is also out of the
| question. Although it would be the best way of measuring it.
|
| --
| Lasse Jensen [fafler at g mail dot com]
| Linux, the choice of a GNU generation.
 >> Stay informed about: Idle powerusage 6200 vs. 8400 

Mr.E Solved! wrote:

> Lasse Jensen wrote:
>> Some time ago i saw an article stating that Geforce 8400 used less power
>> when idle than 6200. I can't seem to find the source anymore. So i
>> brought a no-name 8400 GS a couple of weeks ago. The computer i use the
>> card is always turned on, but rarely used for any gaming. Anyway, the
>> card runs really hot, even when it's not doing anything besides showing a
>> 2D desktop. Is it a possible driver issue? I use the latest binary Linux
>> driver from Nvidia. Does anyone know where i can compare how much power
>> the two card in question should use?
>>
>
>
> Kill-A-Watt
>
>
http://www.amazon.com/P3-International-Kill-Electricity-Monitor/dp/B00009MDBU
>
> or
>
> http://www.thinkgeek.com/homeoffice/gear/7657/
>
> Accurate, cheap, easy to easy, informative, does what you want and more.

It's probably the way to go. It's possible to loan those in public libraries
where i live, so i can pick one up tomorrow.

--
Lasse Jensen [fafler at g mail dot com]
Linux, the choice of a GNU generation.
 >> Stay informed about: Idle powerusage 6200 vs. 8400 

Lasse Jensen wrote:

[regarding the kill-a-watt]

> It's probably the way to go. It's possible to loan those in public libraries
> where i live, so i can pick one up tomorrow.


You can loan framed fine-art reproductions from my local public library
and I thought that unusual and cultural, but your library has geek
gadgets for loan?

Now I feel like I live in a barbarian compound, devoid of any luxury or
progress!
 >> Stay informed about: Idle powerusage 6200 vs. 8400