Kill A Watt

A

annaconda

[H]F Junkie
Joined
Apr 13, 2005
Messages
9,925
Guys if you own "Kill a Watt" please post your Computers Power Usage, by using 3DMARK06.


Specs:

CPU INTEL / AMD @ Speed =
VIDEO CARDs Brand @ Speed =
Total Ram @ Speed =
Total HDD =
Aditional PCI Devices =
Fans + Size =
PSU Watts =
Motherboard =
Water Pump =

TOTAL Watts Idle =
TOTAL Watts Load =
 
CPU @ Speed = E6400 @ 2.67
VIDEO CARDs @ Speed = 450/1320
Total Ram = 2X1gb SIG
Total HDD = 1X250gb
Aditional PCI Devices = X-fi
Fans + Size = 3 120mm fan
PSU Watts = 500watt
etc: SIG

TOTAL Watts Idle = 140watt Load = 160watt
 
CPU @ Speed = 3ghz
VIDEO CARDs @ Speed = 470
Total Ram = 2gb
Total HDD = 5
Aditional PCI Devices = 0
Fans + Size = 2 120mm
PSU Watts = sunbeam nuuo 550

TOTAL Watts Idle = 93 Load = 119
 
So it means PSU does not take 500 watts all the time, even if it rated 500 Watts. Bbq what type of processor do you have ?

Please also state your Video card Brand etc 8800 Series etc.
 
You should specify or standardize the type of load. A load on Super Pi will give significantly lower power draw numbers than a load of 3dmark '06 at 1600x1200 4xAA.
 
J-Mag said:
You should specify or standardize the type of load. A load on Super Pi will give significantly lower power draw numbers than a load of 3dmark '06 at 1600x1200 4xAA.
Click to expand...

Ok how about using 3DMARK06 or INTEL TAT, as long as it takes the whole system to 100%. I think 3DMARK06 is good, so that way, GPU will be under load too.


So 3DMARK06 is the way to go in my Opinion.
 
annaconda said:
So it means PSU does not take 500 watts all the time, even if it rated 500 Watts. Bbq what type of processor do you have ?

Please also state your Video card Brand etc 8800 Series etc.
Click to expand...

Correct. I have an AMD Opteron 146 (stock 200x10 @ 1.4v, oc'd 300x10 @ 1.4v)
eVGA E-Geforce 7800GT (470/1.1)

J-Mag said:
You should specify or standardize the type of load. A load on Super Pi will give significantly lower power draw numbers than a load of 3dmark '06 at 1600x1200 4xAA.
Click to expand...

annaconda said:
Ok how about using 3DMARK06 or INTEL TAT, as long as it takes the whole system to 100%. I think 3DMARK06 is good, so that way, GPU will be under load too.
So 3DMARK06 is the way to go in my Opinion.
Click to expand...

My load was measured using 3dmark 06, while prime95'ing, folding, and transfering 50gb of files back and forth between my drives.
 
CPU INTEL E6300 @ Speed = 3.33
VIDEO CARDs 8800GTS @ Speed = Stock
Total Ram @ Speed = 4GB @ 1000mhz
Total HDD = 2 x 160 SATAII RAID 0
Aditional PCI Devices = ATI 650 HDTV Tunner
Fans + Size = 7 X 120mm Fans
PSU Watts = 500 watts
Motherboard = Evga 680i Sli Motherboard
Water Pump = MCP655

TOTAL Watts Idle = 250 watts
TOTAL Watts Load = 290 watts
 
I think the Kill A Watt is one of the coolest toys I've purchased. I bought 2 - one for work, one for home.

Anyhow, I took some measurements last weekend when I was setting up my new computer. These are from my notes:

System:
e6600 / Intel Bad Axe 2
xfx 7600GT Fatal1ty PCI-e fanless video card
2GB Crucial Ballistix DDR2-1000 memory
WD Raptor 150GB hard drive
2 Yate Loon fans
Lite On DVD drive
Rosewill floppy / CF reader
Corsair HX520 p/s

Base system (2.4GHz):
Idle, 85 watts
SuperPi 1M: 105 watts

2.7 GHz:
Idle, 91 watts
SuperPi 1M: 109 watts
Orthos blend: 128 watts

3.0 GHz
Idle, 100 watts
SuperPi 1M, 120 watts
Orthos blend: 155 watts

I'm still tweaking from there. I'm not going for an aggressive overclock, just something that modestly blends overclocking, power consumption, and thermal output. I'm currently at 3015 MHz and very stable, with orthos blend 2-hour temperature at 48-50C.

FWIW, my MythTV box, based on a cheap ECS motherboard, Athlon 2800, 1GB OCZ RAM, and Seagate 250 GB IDE hard drive, draws 75 watts while playing.
 
Specs: The build is outlined here with pictures.

CPU INTEL / AMD @ Speed = E6700 @ 3GHz
VIDEO CARDs Brand @ Speed = BFGTech 8800GTX and 6600GT
Total Ram @ Speed = 2GB @ 1000MHz
Total HDD = 8 total
Aditional PCI Devices = RAID Card
Fans + Size = 2 @ 120 1 @ 90
PSU Watts = 600w
Motherboard = Asus Commando P965

TOTAL Watts Idle = 296
TOTAL Watts Load = 410

This load was taken running Orthos and RTHD demo full screen on the 30.
 
Hate to burst your bubble, guys, but Kill-a-Watt is not very accurate.

Mythbusters tested light bulb power consumption a few months ago. One of the bulbs they tested was an LED bulb, with 172 separate LED's on it. They estimated, based on the KaW's readings, that it drew around 1W total. I can't find it now, but my dad and I were both intrigued by their results. We tracked down the bulb to its manufacturer: 10 watts!

I think KaW can only measure resistive load, not capacitive. Your electric company bills you on TOTAL consumption, not resistive alone.
 
My Monster Server :D

CPU INTEL / AMD @ Speed = P4 @ 1.7ghz
VIDEO CARDs Brand @ Speed = Intel 845G
Total Ram @ Speed = 1GB PC3200
Total HDD = 1 (2.5" 40GB 5400rpm laptop hard drive)
Aditional PCI Devices = none
Fans + Size = none
PSU Watts = 500w
Motherboard = LV-670 Mini ITX Motherboard

TOTAL Watts Idle = 30
TOTAL Watts Load = 77 (and that is because P4 1.7ghz rated 70watts) :D

Running: Windows 2003 Enterprise Edition & Windows 2003 Exchange.
DNS.
Active Directory,
IIS.
SSH
File Server.
VPN.
 
Pfft, my server is even better.

P4 1.5 Willamette
Intel 845 chipset
256mb sdram
a 40gb 7200rpm drive (thanks khanable <3)
fsp 300w psu

idle: 28w
load: 60w

Of course, all it's doing is running ratbox, ratbox_services, irssi, and gnu screen on FreeBSD 6.2_release.
 
mavalpha said:
I think KaW can only measure resistive load, not capacitive. Your electric company bills you on TOTAL consumption, not resistive alone.
Click to expand...

Actually, the Kill-A-Watt does also measure power factor, so I believe it would then have to know the difference between a resistive and capacitive load. If it only worked accurately on resistive loads, it would only work with devices with a power factor of 1 (apparent power = real power.)
 
You're probably right, Jonny, but the fact of the matter is that something is still fishy. I'm not the only one that noticed, thank goodness!

Read Excedrin's post: http://reddit.com/info/15km8/comments/c15nen
He linked the wrong bulb (he points to a 138-LED, but the MB version was 172), but that should- if anything- return a lower power consumption. The manufacturer claims 10W on the smaller version. He heard the same 1W report from Adam and Jamie as I did, giving a minimum 900&#37; margin of error! :eek:
 
mavalpha said:
You're probably right, Jonny, but the fact of the matter is that something is still fishy. I'm not the only one that noticed, thank goodness!

Read Excedrin's post: http://reddit.com/info/15km8/comments/c15nen
He linked the wrong bulb (he points to a 138-LED, but the MB version was 172), but that should- if anything- return a lower power consumption. The manufacturer claims 10W on the smaller version. He heard the same 1W report from Adam and Jamie as I did, giving a minimum 900% margin of error! :eek:
Click to expand...

1 and 10 sounds much like a error somewhere. Call the company and find out, perhaps it's a typo.
 
mavalpha said:
You're probably right, Jonny, but the fact of the matter is that something is still fishy. I'm not the only one that noticed, thank goodness!

Read Excedrin's post: http://reddit.com/info/15km8/comments/c15nen
He linked the wrong bulb (he points to a 138-LED, but the MB version was 172), but that should- if anything- return a lower power consumption. The manufacturer claims 10W on the smaller version. He heard the same 1W report from Adam and Jamie as I did, giving a minimum 900% margin of error! :eek:
Click to expand...

Extremely doubtful to beyond realistic given that when a known load is running through the Kill-A-Watt it measures it far more accurately than the scenario you pose.

After watching that show over the last couple of years I would have to say the error lies with the Mythbusters, used to be interesting until the wheels came off.
 
I dont have one but my neighbor does, so I borrowed his. The results explain why my electric bill is high!

Specs:

CPU INTEL @ Speed = 3.4
VIDEO CARD 2 8800GTX @ Speed = 600
Total Ram 4GB Corsair Dominator @ Speed = 1066
Total HDD = 8
Aditional PCI Devices = 1 SATA card
Fans + Size = 12x120/2x90
PSU Watts = 1Kw
Motherboard = eVGA 680i
Water Pump = 2 Aqua

TOTAL Watts Idle = 352
TOTAL Watts Load = 501 3DMark 06
TOTAL Watts Load = 519 COD2 Max Settings 19x12
TOTAL Watts Load = 527 COH Max Settings 19x12

You should see the numbers from audio my system! Monoblocks are very hungry.
 
CPU INTEL Core 2 Duo E4300 @ Speed = 1.8Ghz
VIDEO CARDs 7600GS @ Speed = 400Mhz
Total Ram @ Speed = 1Gig/800Mhz
Total HDD = 7HD
Aditional PCI Devices = PCI IDE card
Fans + Size = about 6
PSU Watts = 300 Seasonic
Motherboard = cheapo Socket 775
Water Pump = None

TOTAL Watts Idle = 100 watts
TOTAL Watts Load = 130 watts (glxgears + Bonnie++)
 
unhappy_mage said:
Jeez, either that 12% overclock Kyle's got going makes the chip draw a lot more, or the 8800 is friggin heavy on 12V rails. The TDPs on these chips is the same, Kyle's got an 8800 more than longblock454 and a 680i instead of a generic chipset, and everything else is sort of the same. There's a 280W difference between these two systems at load :eek:
Click to expand...

And a 6600GT etc.....
 
Paul_Johnson said:
And a 6600GT etc.....
Click to expand...

Yeah, but I was guesstimating that the 7600GS and the 6600GT would be approximately equivalent in power draw. 131W (for the whole system) versus 152W at extreme's power calculator, in that order. So the 6600 is even heavier than the 7600. They peg the 8800GTX at 109W.

In other words, holy cow the 680 takes a lot of power. Unless extreme's numbers are way off.
 
unhappy_mage said:
Yeah, but I was guesstimating that the 7600GS and the 6600GT would be approximately equivalent in power draw. 131W (for the whole system) versus 152W at extreme's power calculator, in that order. So the 6600 is even heavier than the 7600. They peg the 8800GTX at 109W.

In other words, holy cow the 680 takes a lot of power. Unless extreme's numbers are way off.
Click to expand...

I am not sure, but I believe the 680i consumes as much power as my old Tyan 2895 did and possibly more.
 
annaconda said:
Guys if you own "Kill a Watt" please post your Computers Power Usage, by using 3DMARK06.
Click to expand...


Specs:

CPU INTEL / AMD @ Speed
Intel QX6700 quad core
VIDEO CARDs Brand @ Speed
single 8800GTS pny @ stock speed
Total Ram @ Speed
8 GB DDR2-667
Total HDD
6 x 500GB sata2 MaxtorMaxline, 1 x Pata 7200.10 Seagate
Addtional PCI Devices
Strangely, none
Fans + Size
14 x 60mm (2 per hdd) 4 x 120mm (2 inflows, 2 outflows), 4 x 40mm (interstitials creating airflow where there isn't any), 2 x 80mm (2 for extra psu area outflow), 1 x 92mm on cpu, one in psu, one on video card
PSU Watts
PCP&C 750W Silencer
Motherboard
Intel DG965WH
Water Pump
None
TOTAL Watts Idle
232. Confirmed by my UPS within 5-10 pts.
TOTAL Watts Load
310. Again, confirmed by UPS.


LCD's are NOT on the killawatt / ups for when I did this testing. But they normally are. When on, the usage jumps to over 400, on my 1500VA ups. That of course isn't affecting my PCP&C 750, but those #'s still affect my electric bill. Westy pulls 170W or so.
Why so many fans? Fans are cheap, are proven to work as a cooling method, thus making them probably the MOST cost effective way to protect hardware from heat damage, instead of exotic LN2, peltier, or whatever solutions., and I've got the room for them. The combined force of ALL these
fans still doesn't make a hum I can hear.

You take a 7200RPM sata2 hdd, and you blow two 60mm fans at it's bottom, and a 120mm at it's side (i.e. top and bottom) it's gonna stay cool. My hdd's run in the high 20s, to low 30s, and they WORK hard.
 
CyberDeus Your readings are amazing with Quad Core, 8800gts, plus all those fans and hard drives.

I am confused either your PSU is giving wrong numbers or my Kill A Watts is not correct. I have higher numbers than you, with e6300, 8800gts, 2HDDs and only 6 fans. :confused:
 
annaconda said:
CyberDeus Your readings are amazing with Quad Core, 8800gts, plus all those fans and hard drives.

I am confused either your PSU is giving wrong numbers or my Kill A Watts is not correct. I have higher numbers than you, with e6300, 8800gts, 2HDDs and only 6 fans. :confused:
Click to expand...

You are overclocked, no, annaconda?
As I said, I've double checked the draw with my UPS, as well. There are people here with 91 watts idle reported, and not THAT far below my config, 91 watts seems amazing to me..

If you look @ Kyle's 296/410 by comparison, you will see what overclocking does to your power consumption, by all accounts, Stock, I should pull more than he does.
Remember, NOTHING in my system is overclocked (at least, system1, that is).. I never overclock my primary system. I only overclock the systems I can afford to have go up in flames.
 
CPU INTEL Dothen (Pentium M)@ Speed =2.0
VIDEO CARDs nvidia 7800go gtx @ Speed =470x1300
Total Ram 2gb @ 533
Total HDD =60gb 7200
PSU Watts =130w power adapter
Motherboard =dell xps2 laptop


TOTAL Watts Idle =44
TOTAL Watts Load =123
 
Asus L1N64-SLI
Two FX74 CPU's
4GB DDR2-1000 (4 x 1GB sticks)
Two eVGA 8800GTX
Two Seagate Barracude 7200.10 750GB drives
BenQ DVD640
Ultra M998 case w/ X3 1000W PSU

From the wall: 636.9W
 
it looks like the Seasonic uses the least amount of wattage. is that true?

thanks
 
elyrly said:
it looks like the Seasonic uses the least amount of wattage. is that true?
Click to expand...

How do you figure?

A power supply only puts out what's asked of it. So output wattages and draw from the wall are going to be different per build. Every one of these builds are different from the next, so you can't make a linear comparison.
 
but you also need to consider the efficiency of the PSU. most range from 75%-80%
 
These days, most decent ones are 80% and up.

I guess you're saying "looks like the Seasonic uses the least amount of wattage" based on what you've heard and not based on anything you're reading in this thread (because most of the posts in this thread don't state their PSU.)

Seasonics are efficient, but it is not the most efficient power supplies brand any more.
 
You must log in or register to reply here.
Back
Top