got a power meter? what is your watts/TB?

In regards to measuring power consumption, how accurate (compared to Kill-a-watt or UPS reading) do you guys think it is to put a 1ohm resistor in series with the neutral power wire then measure the voltage drop across the resistor and solve for power consumed by the computer?

I made a custom power cable with a 1ohm resistor, and through this method I am getting 45watts on my Celeron 430 + WD SE16 640GB + WD Green 1TB WHS. 27 W/TB
 
Advertised Capacity (TB): 16.5
Wall Plug Power Draw @ v (idle) (W): 129
W/TB: 7.8
Measurement: UPS

Motherboard: P6T Deluxe v.2
CPU: i7 920 @ stock
Graphics: GTX 285 oc'd
Controller: AOC-SASLP-MV8
Hard Drives: 8x WD20EADS + SSDSA2M160G2R5 + WD3000HLFS
PSU: HX1000W
Other: 2209WA (included)
 
Last edited:
In regards to measuring power consumption, how accurate (compared to Kill-a-watt or UPS reading) do you guys think it is to put a 1ohm resistor in series with the neutral power wire then measure the voltage drop across the resistor and solve for power consumed by the computer?

I made a custom power cable with a 1ohm resistor, and through this method I am getting 45watts on my Celeron 430 + WD SE16 640GB + WD Green 1TB WHS. 27 W/TB

Please don't do that. That is dangerous. We're not talking PC power supply voltage here.

Also, it's not even accurate. There's no way to know the power factor of your PS, actual current you will measure through your resistor will be higher then the value needed to calculate power consump.
 
Advertised Capacity (TB) - 12.6TB
Wall Plug Power Draw @ v (idle) (W) - ~165W
W/TB - 13W/TB
Measurement - APC UPS LCD

Motherboard - Asus P6T6 WS Revolution
CPU - Intel Xeon W3520
Graphics - nVIDIA 7300GT
Controller
-- LSI SAS 3081E-R
-- Intel ICH10R

Hard Drives
6 x WD 1TB Green SATA
5 x Hitachi 1TB SATA
4 x Seagate 320GB SATA
1 x WD 320GB SATA

PSU - Antec TruPower 750W
Other
 
Here's my main.



Advertised Capacity (TB) --2.564 TB
Wall Plug Power Draw @ v (idle) (W) --120W
W/TB -- 21.4
Measurement -- UPS

Motherboard -- Asus M4A79
CPU -- AMD Phenom II X6 1090T
Graphics -- AMD/ATI Radeon 5870
Controller -- AMD SB750
Hard Drives -- Patriot Torqx-64GB, Western Digital Black 1TB and Green 1.5TB
PSU -- Silverstone Strider 750w
 
Last edited:
Advertised Capacity (TB) 11 tb
Wall Plug Power Draw @ v (idle) (W) 146w
W/TB 13.27
Measurement Kill A Watt

Motherboard Tyan S5162
CPU Pentium D 930
Graphics integrated
Controller OS=onboard SATA, everything else on Supermicro AOC-SAT2-MV8
Hard Drives
WD10EACS 1tb OS
WD10EACS 1tb pool
WD10EACS 1tb pool
WD10EACS 1tb pool
WD10EACS 1tb pool
WD10EADS 1tb pool
ST315003 41AS 1.5tb pool
ST315003 41AS 1.5tb pool
ST320003 42AS 2tb pool
=======================
10,243gb formatted

PSU Antec Earthwatts EA-650
Other 2x1gb RAM
4x120mm fans
DVD drive

The Gigabyte motherboard I have had in my WHS went out a few weeks ago and I substituted an ECS that I had but was having issues with it (this setup was with an E2200 cpu). I saw the Tyan board on sale at Newegg the day after Xmas for $35 and had a Pentium D 930 laying around so I ordered it and the RAM. I can now run the Supermicro card at full speed, it's obvious that the transfer of data back into the pool is only going to take about 1/5th the time. Having a real server board definitely feels much better overall and my overall wattage is only up 16w despite the older P4 architecture.

I did notice an interesting thing while connecting my 1tb EACS drives. Hooked into the mobo SATA ports each one only added 3w. Hooking them into the Supermicro card added 6w each overall (3w for drive presumably and 3w for the controller).

Running just the board, 930, ram, dvd, 2x120mm fans, and one 1tb EACS drive I was showing 93w at idle. The SM card added 5w with no drives connected.
 
Advertised Capacity (TB) 20.2
Wall Plug Power Draw (idle) (W) 128
W/TB 6.3
Wall Plug Power Draw (idle - RAID drives spun down) (W) 93
W/TB 4.6
Measurement Kill-a-Watt

Motherboard: IEI KINO-780AM2
CPU: AMD Athlon 400e
Controller: Areca 1260
Hard Drives:
1x Maxtor 200GB 7200RPM SATA
10x WD WD20EADS 2TB
PSU: Antec EA500
Other:
2x Supermicro CSE-M35T
2x 120mm fan
1x 92mm fan
USB DVD drive
 
Last edited:
have updated the list
as of today (2010.06.08) the list SHOULD be up-to-date
 
What does the watts/TB metric mean, anyway? Anyone can cram 8 x 2 TB drives on a board in a JBOD config with a Celeron DC and claim their watts/TB is greater than my 500 GB RAID-1 OS and 8 x 2 TB RAID-10 storage config using a Xeon X5560.

It says nothing meaningful if you're disregarding performance, reliability, redundancy, practicality, etc. At least, I think. :p
 
I can get plenty of performance out of a low power file server, it's based on need just like the rest of the system. If low power is more important obviously performance isn't top of the list.

A WD Green drive can be just as reliable as any other HDD, a well build ATOM based SM system is just as reliable as it's bigger SM parts.

Redundancy is over rated, I'd take a good set of back ups over a redundant RAID system any day.

For someone who just needs storage... a low power system makes WAY more sense then a huge quad server with unessassary power.. I don't think my system would transfer my files much faster with a Xeon then a low power option.
 
I started this thread as i had the misconception that an atom based WHS would have the lowest wallplug-power-draw. I wanted to find out if this opinion was correct.

Hopefully the thread will also prove useful to those, who like me, want to build a medium capacity, low-wallplug-power draw server.
For myself, I learnt that my second WHS build (in progress) could be based on an core i3 with no detrimental affect upon power draw. For reference, the motherboard is a Supermicro X8SIL-F

If this thread is not useful if will die away quietly
ZZZzzzzz.....

What does the watts/TB metric mean, anyway? Anyone can cram 8 x 2 TB drives on a board in a JBOD config with a Celeron DC and claim their watts/TB is greater than my 500 GB RAID-1 OS and 8 x 2 TB RAID-10 storage config using a Xeon X5560.

They can indeed, and the number would prove a useful addition to the list
The watts/TB metric is based on installed capacity, not how that capacity is utilised.
i.e. your storage would be 17Gig rather than the available 8.5gig

It says nothing meaningful if you're disregarding performance, reliability, redundancy, practicality, etc. At least, I think. :p

The watts/TB metric has little or nothing to do with reliability, redundancy, practicality, etc.
Reliability is a measure of the components used, and those components need not have a high power consumption
A WHS server with folder duplication enabled with provide a high degree of redundancy
Practicality is based on one's needs.

Hopefully that answers your questions as to the whys and wherefores of this thread
 
Last edited:
I built a Mini ITX server for exactly this reason, I wanted something always on, running W2008 Server R2, with a lot of storage that I could access from anywhere and host some personal websites on through IIS7.

It's been awesome so far!

6TB 3x2TB WD drives
Zotac MiniITX C2D E3500
4GB Ram
500w PSU

Draws 43w with a kill-a-watt meter attached. Thats about double my Dlink DNS-323 however its infinitely more usable for me as it hosts websites, virtual machines, VPN, etc.
 
What does the watts/TB metric mean, anyway? Anyone can cram 8 x 2 TB drives on a board in a JBOD config with a Celeron DC and claim their watts/TB is greater than my 500 GB RAID-1 OS and 8 x 2 TB RAID-10 storage config using a Xeon X5560.

It says nothing meaningful if you're disregarding performance, reliability, redundancy, practicality, etc. At least, I think. :p

I think I pointed the same out in the past. I believe my comments went along the lines of start with one hard drive and add more. The watts/TB go down.

Adding a removable hard drive to a system adds as much storage space as one can afford to buy without increasing the power requirements.

But look at the guy who can afford to buy 100TB of drives. He has no concept that there are much cheaper solutions to his problem. People are not always bright.
 
It's not like this is an ordered list of watts/TB, right? Or ordered by cost, performance, etc. It's just a poll to find out what people have and what kind of power usage they are seeing, with the hardware they posted.
 
But look at the guy who can afford to buy 100TB of drives. He has no concept that there are much cheaper solutions to his problem. People are not always bright.
Huh? What cheaper solution did I have available?
 
Oh. Funny thing is, that's almost exactly what I've built. What would've been cheaper?
 
Adding a removable hard drive to a system adds as much storage space as one can afford to buy without increasing the power requirements.

unless you know of a magical external hard drive, then adding an external harddrive will increase the power requirements of the server; merely indirectly.

croakz is correct, the list is merely a list of the hardware used and power obtained. the engineering geek (i am engineer by training) is fasinated to know what builds people have and the power the builds use. :cool:
 
Just Bought myself a rebranded Kill-a-Watt

this rig is an all in one LAN box hence the high end CPU, GPU ect otherwise i am sure i could get much lower numbers

Butcher9_9

Advertised Capacity (TB) 28.48TB
Wall Plug Power Draw @ v (idle) (W) 300W (this is including a UPS and monitor on standby)
W/TB10.53W
Measurement Rebranded Kill-a-Watt


Motherboard Gigabyte X58 UD4P
CPU I7 920
Graphics ATI 5870
Controller Adaptec 51645
Hard Drives 6X 2tb 5400rpm, 2X2tb 7200rpm, 8X1.5tb 5400rpm, 128gb SSD and 7200rpm 2.5" 320gb
PSU Silverstone Strider 850W
Other Intel ET dual port network card, 8X 120mm fans+ controller, see sig

the biggest part of my power consumption is my monitors which each draw 150w (not included in figure above
 
unless you know of a magical external hard drive, then adding an external harddrive will increase the power requirements of the server; merely indirectly.

i think he means that when a removable HDD is removed it draws no power, i use a large number of drives in hot swap bays as my data backup for this reason

the HDDs are safer in a box than a PC (power surges ect) and they draw 0 watts
 
rlisms

Advertised Capacity (TB) 10TB
Wall Plug Power Draw @ v (idle) (W) 47, 35 w/ disks idle
W/TB 4.7, 3.5 w/ disks idle
Measurement Kill-a-Watt

Motherboard Gigabyte GA-785GM-US2H
CPU AMD Athlon II X2 240 @1.175V
Graphics Integrated @0.975V
Controller Integrated
Hard Drives 5 x 2TB SAMSUNG Spinpoint F3EG
PSU Rosewill RG530-S12 530W
Other 2x 120mm Scythe Kama Flow 2 @900rpm

More info and pics in the 10TB+ thread HERE
 
I have both a kill - a - watt and a UPS the UPS says 220 watts and the kill-a-watt says 231 watts. Thats a pretty big difference so I'm not sure which one to go by. The PC is plugged into the Kill-A-Watt that is plugged into the UPS.

This is for my quadcore PC with 2 video cards and 7 hard drives about 5.2TB
I haven't measured my server yet because I'm going to pull 2 of the drives fromt his PC and move them over there.


Advertised Capacity (TB) 5.2
Wall Plug Power Draw (idle) (W) 230
W/TB 44.23
 
The Kill-A-Watt claims 0.2% accuracy. What accuracy does your UPS claim? Is it power-factor corrected?
 
I have both a kill - a - watt and a UPS the UPS says 220 watts and the kill-a-watt says 231 watts. Thats a pretty big difference so I'm not sure which one to go by. The PC is plugged into the Kill-A-Watt that is plugged into the UPS.

Your UPS draws 11 watts (assuming both are 100% accurate)?

How about plugging the Kill-A-Watt into the wall, then the UPS, then to PC? Do both read ~231 watts?
 
Intel "Suitcase" NAS

Advertised Capacity (TB) 4TB
Wall Plug Power Draw @ v (idle) (W) 62watts idle (disks spinning) after spindown it's around 30-40 watts.
W/TB 15.5
Measurement KILL-A-WATT

Motherboard INTEL Nas board (SS4200-E)
CPU Celeron 430 1.6GHz
Graphics none
Controller onboard
Hard Drives Two seagate 7200.11 and Two seagate 7200.12 (all four Drives in raid5)
PSU 250watt Delta
Other full product info here:
http://www.intel.com/support/motherboards/server/ss4200-e/sb/CS-028562.htm
 
Last edited:
Back
Top