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

Asposium

Limp Gawd
Joined
Mar 13, 2009
Messages
351
Seeing people in the >10TB thread posting their server power draw got me wondering...

For those with a power meter, what is the watts/TB value for your server?
Only included storage included in the power measurement.

_______________

Last Modified 2010.06.19




Butcher9_9
Added 2010.06.19

Advertised Capacity 28.48
Wall Plug Power Draw @ 220v (idle) (W) 300
W/TB 10.53
Measurement kill-a-watt (UK version)

Motherboard Gigabyte X58 UD4P
CPU Intel i7 920
Graphics ATI 5870
Controller Adaptec 51645
Hard Drives
• 1x 128GB SSD
• 1x 320GB 7200rpm 2.5inch
• 6x 2TB 5400rpm
• 2x 2TB 7200rpm
• 8x 1.5TB 5400rpm
• 10x WD WD20EADS 2TB
PSU Silverstone Strider 850W
Other None
• Intel ET dual NIC
• 8x 120mm fan

_______________

Croakz
Added 2010.06.08

Advertised Capacity 20.2
Wall Plug Power Draw @ 220v (idle) (W) 128
W/TB 6.3
Measurement kill-a-watt (UK version)

Motherboard IEI KINO-780AM2
CPU AMD Athlon 400e
Graphics ?
Controller Areca 1260
Hard Drives
• 1x Maxtor 200GB 7200RPM SATA
• 10x WD WD20EADS 2TB
PSU Antec EA500
Other None
• 2x 120mm fan
• 1x 92mm fan

_______________

[LYL]Homer
Added 2010.05.01

Advertised Capacity (TB) 11
Wall Plug Power Draw @ v (idle) (W) 146
W/TB 13.27
Measurement kill-a-watt

Motherboard Tyan S5162
CPU Intel Pentium D 930
Graphics On-board
Controller
• On-board
• 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
PSU Antec Earthwatts EA-650
Other
• 4x120mm Fans

_______________

SicKlown42012
Added 2009.12.26

Advertised Capacity (TB) 1.704
Wall Plug Power Draw @ v (idle) (W) 120
W/TB 70.4
Measurement UPS

Motherboard Asus M4A79
CPU AMD Phenom II X4 940BE
Graphics AMD/ATI Radeon 5870
Controller AMD SB750
Hard Drives
• Patriot Torqx-64GB
• WD Black 640GB
• WD Green 1TB
PSU Silverstone Strider 750W
Other

_______________

megabit
Added 2009.12.26

Advertised Capacity (TB) 12.6
Wall Plug Power Draw @ v (idle) (W) 165
W/TB 13
Measurement UPS

Motherboard Asus P6T6 WS Revolution
CPU Intel Xeon W3520
Graphics nVIDIA 7300GT
Controller
• On-board
• LSI SAS 3081E-R
Hard Drives
• 6x WD 1TB Green SATA
• 5x Hitachi 1TB SATA
• 4x Seagate 320GB SATA
• 1x WD 320GB SATA
PSU Antec TruPower 750W
Other

_______________

quixotic
Added 2009.12.26

Advertised Capacity (TB) 16.5
Wall Plug Power Draw @ v (idle) (W) 129
W/TB 7.8
Measurement UPS

Motherboard Asus P6T Deluxe v2
CPU Intel i7 920
Graphics GTX 285 oc'd
Controller Supermicro AOC-SASLP-MV8
Hard Drives
• 8x WD20EADS
• SSDSA2M160G2R5
• WD3000HLFS
PSU HX1000W
Other

_______________

Zepher
Added 2009.12.25

Advertised Capacity (TB) 7
Wall Plug Power Draw @ v (idle) (W) 280
W/TB 40
Measurement ?

Motherboard ?
CPU Intel Q6600 @ 3.02GHz
Graphics eVGA GTX 285
Controller
Hard Drives
• 3x WD 1TB Black
• 2x WD 500GB My Books
• 1x Seagate 1.5TB
• 2x WD 250GB
• 1x Seagate 400GB
• 1x WD 500GB
• 1x Seagate 320GB
PSU Antec True Power New 650W
Other

_______________

okashira
Added 2009.12.25

Advertised Capacity (TB) 6
Wall Plug Power Draw @ v (idle) (W) 30
W/TB 5
Measurement UPS

Motherboard Intel S4200-E
CPU Intel S4200-E
Graphics On-board
Controller On-board (ICH7)
Hard Drives 4x WD Green 1.5TB
PSU Stock Intel S4200-E
Other

_______________

Asposium
Added 2009.12.04

Advertised Capacity 7.32
Wall Plug Power Draw @ 220v (idle) (W) 82
W/TB 11.2
Measurement kill-a-watt (UK version)

Motherboard Intel D945GCLF2D
CPU Intel Atom 330
Graphics On-board
Controller
• On-board
• Adaptec 1420SA
Hard Drives
• 1x Seagate 7200.11 320GB
• 3x Seagate 7200.12 1TB
• 2x Seagate 5900.LP 2TB
PSU Corsair HX420
Other None

_______________


MixManSC
Added 2009.10.28

Advertised Capacity ?
Wall Plug Power Draw @ 220v (idle) (W) 190
W/TB ?
Measurement ?

Motherboard Intel D945GCLF2D
CPU Intel Dual Core Xeon
Graphics 8800GTS
Controller ?
Hard Drives
• 1x Intel G2
• 4x 15000 SAS
PSU 1kW
Other 32GB FB-DIMM

_______________


Yakyb
Added 2009.10.28

Advertised Capacity 8
Wall Plug Power Draw @ 220v (idle) (W) 103
W/TB 12.87
Measurement Kill-a-watt

Motherboard MSi P55 GD65
CPU Intel Core i7 860
Graphics ATI 3450
Controller On-board
Hard Drives
• 1x Seagate 500GB
• 5x WD GB
PSU Antec Signature 650w
Other Server 2008 R2

_______________


alan.p
Added 2009.09.04

Advertised Capacity 3
Wall Plug Power Draw @ 220v (idle) (W) 75
W/TB 25
Measurement Kill-a-watt

Motherboard ?
CPU AMD AM2-1600LE
Graphics On-board
Controller On-board
Hard Drives 3x Samsung 1TB
PSU ?
Other ?

_______________


sniggle
Added 2009.09.03

Advertised Capacity 4.572
Wall Plug Power Draw @ 220v (idle) (W) 217
W/TB 47.46
Measurement 1300VA APC UPS

Motherboard Asus P5WDH
CPU Q6600 O/C to 3GHz @ 1.31v
Graphics eVGA GeForce 8600GTS
Controller Highpoint RocketRaid 2320
Hard Drives
• 2x WD Raptor 36GB RAID1 (os)
• 6x WD RE2 750GB RAID5
PSU Corsair 750HX
Other ?

_______________


Adidas
Added 2009.08.24

Advertised Capacity 4.5
Wall Plug Power Draw @ 220v (idle) (W) 230
W/TB 51.11
Measurement Kill-a-watt

Motherboard Biostar G33 mATX
CPU Intel Celeron E1400
Graphics On-board G33
Controller Highpoint RocketRaid 2300, 2200
Hard Drives 7x WD 750AAKS 750GB
PSU Antec 650W
Other Norco 4020

_______________


astroidea
Added 2009.08.24

Advertised Capacity 5.79
Wall Plug Power Draw @ 220v (idle) (W) 65
W/TB 11.22
Measurement Kill-a-watt

Motherboard GA-945GCM-S2C
CPU Intel Celeron 440 2.0GHz
Graphics On-board
Controller Supermicro AOC-SASLP-MV8
Hard Drives
• 2x Seagate 1.5TB
• 2x WD 640GB
• 2x Hitachi/Samsung 1TB
• 1x Maxtor 80GB
PSU Antec Earthwatts 450w
Other ?

_______________


houkouonchi
Added 2009.08.24

Advertised Capacity 20
Wall Plug Power Draw @ 220v (idle) (W) 350
W/TB 17.5
Measurement Watts-up-pro

Motherboard MSI nforce 750i SLI board
CPU Intel Core 2 Quad Q6600 2.4GHz
Graphics Nvidia Geforce GTX 260
Controller Areca ARC-1280ML
Hard Drives 20x Seagate 7200.? 1TB
PSU Corsair 850w
Other
• Norco 4020
• Creative Audigy 2 ZS
• 8GB DDR2-1000

_______________


DeChache
Added 2009.08.23

Advertised Capacity 2.75
Wall Plug Power Draw @ 220v (idle) (W) 55
W/TB 20
Measurement APC UPS

Motherboard Gigabyte P35 DS3R
CPU Intel Core 2 E8400
Graphics ATI Rage 6 PCI
Controller On-board
Hard Drives
• 2x WD GP 1TB
• 1x Maxtor 500GB
• 1x WD 350GB
PSU Antec Earthwatts 380
Other ?

_______________


kharan5876
Added 2009.08.21

Advertised Capacity (TB) 15.1
Wall Plug Power Draw @ 110V (idle) (W) 140
W/TB 9.3
Measurement Kill a watt

Motherboard Supermicro X7SBA w/ IPMI add on card
CPU Intel Core 2 Q9450
Graphics Onboard, XGI Volari
Controller
LSI SAS 3801E Pci-E Host Bus Adapter
Chenbro ck13601 Sas Expander
Hard Drives
• 10x Samsung Spinpoint F2EG HD154UI 1.5TB 5400
• 1x 100GB 2.5" 5400RPM drive
PSU Corsair HX850
Other
• Intel Quad port PCI-X Gigabit NIC
• 5x 80mm case fans (three limited to 7V)

_______________

Dan_D
Added 2009.08.20

Advertised Capacity (TB) 1.16
Wall Plug Power Draw @ v (idle) (W) 450
W/TB 388
Measurement UPS

Motherboard EVGA X58 3X SLI Classified
CPU Intel Core i7 920 @ 4.2GHz
Graphics 3x BFG Technologies Geforce GTX 280 OC (3-Way SLI)
Controller ICH10R
Hard Drives
• 2x Western Digital Black 500GB
• 2x Intel X25-M 80GB
PSU Thermaltake ToughPower 1200watt
Other

_______________

Asposium
Added 2009.06.26

Advertised Capacity 4.32
Wall Plug Power Draw @ 220v (idle) (W) 75
W/TB 17.4
Measurement kill-a-watt (UK version)

Motherboard Intel D945GCLF2D
CPU Intel Atom 330
Graphics On-board
Controller
• On-board
• Adaptec 1420SA
Hard Drives
• 1x Seagate 7200.11 320GB
• 4x Seagate 7200.12 1TB
PSU Corsair HX420
Other None

_______________

Asgards
Added 2009.06.30

Advertised Capacity 19
Wall Plug Power Draw @ 110v (idle) (W) 135
W/TB 7.0
Measurement ?

Motherboard



GigaByte GA-MA74GM-S2

CPU ?
Graphics ?
Controller
• 1x Supermicro AOC-SAT2-MV8
• 1x Sil3114
Hard Drives
• 1x Western Digital WD10EADS
• 12 x Segate ST31500341AS
PSU ?
Other ?

_______________

Asgards
Added 2009.06.30

Advertised Capacity (TB) 13.5
Wall Plug Power Draw @ 110v (idle) (W) 120
W/TB 8.88
Measurement ?

Motherboard GigaByte GA-MA74GM-S2
CPU ?
Graphics ?
Controller
• 1x Supermicro AOC-SAT2-MV8
• 1x Sil3114 /
Hard Drives
• 1x WD10EADS
• 2x HD103UJ
• 6x Seagate 7200.11 1500TB (ST31500341AS)
• 2x HD753LJ
PSU ?
Other ?

_______________

Template

Advertised Capacity (TB)
Wall Plug Power Draw @ v (idle) (W)
W/TB
Measurement

Motherboard
CPU
Graphics
Controller
Hard Drives
PSU
Other

____________________
 
Last edited:
power was measured with the UK version of the kill-a-watt meter
so, 75w is wall plug power draw
surely if power usage is measured as wall plug draw i need not be concerned with power factor and VA???
 
Depending on your power supply efficiency you could be off by 30%.

With anything inductive, such as a computer PS, watts doesn't really mean much. Take for example why they rate UPSs in VAs?

What does the kill-a-watt say the VAs are?
 
Depending on your power supply efficiency you could be off by 30%.

With anything inductive, such as a computer PS, watts doesn't really mean much. Take for example why they rate UPSs in VAs?

What does the kill-a-watt say the VAs are?

While this is certainly true, the OP is accurately measuring the power being pulled from the socket. Whether the electricity is used by an electronic component or turned into heat by an inefficient power supply, this is the number that effects their power bill.
 
My UPS (APC 1500VA LCD) is saying that my server is currently using around 120-130W, and at the moment, I'm adding two more drives to my array, so I suppose that's safe to say it's the reading I would get under load conditions.

9 x ST31500341AS with Corsair CMPSU-450TX 450W PSU equalling about 9.779TB (1.397TB x 7 minus formatted after two drives subtracted for RAID 6 parity) makes my watts/TB about 12.78.

Not sure if my UPS is the most accurate measure of the wattage being pulled by the server, but nothing else is connected to it (other than a monitor that is used sparingly), so I suppose a Kill-A-Watt would be more accurate.
 
This is a good idea

my 3.5 TB system currently pulls 170W

putting me at 48.5 W/TB

i will hope to upgrade that soon tho as currently there are loads of 400/500/750 GB drives there
 
Array size : 2.58TB (4x1TB WD10EADS in RAID5 with RAID1 for system)
Powr usage at idle : 45W at socket
17.44W/TB

System:
Asus M3N78, AMD X2 4850e, 4xWD10EADS, Modu82-425W

PS: It doesn't take much more power when used, max CPU usage is typically 10% because video is accelerated via VDPAU in XBMC, Samba is not a bnig CPU hog...
 
interesting thread... i think it would be helpful to list some of the components... i've thought about checking what my APC backups says with the P3 kill a watt and the Seasonic Power Angel.
and are we doing Advertised TB? or actual TB? 1TB advertised? or the ~930GB you actually get? I'll take a look and list what my APC backups says and cross check w/the watt meter later...
 
HP EX475

Seagate 750GB
WD GP 640GB
2 x WD GP 1TB

3.157GB actual storage.

60 Watts at idle = 19W/TB
 
and are we doing Advertised TB? or actual TB? 1TB advertised? or the ~930GB you actually get? I'll take a look and list what my APC backups says and cross check w/the watt meter later...

Advertised TB

as the wall plug power draw has little to do with how the drive are arranged

by that i mean

10 drives as JBOD will have approx the same draw as those same 10 drives configured as RAID 6
 
Whether the electricity is used by an electronic component or turned into heat by an inefficient power supply, this is the number that effects their power bill.

True, for residential, often not the case for industry/commercial.

Since the majority of users will measure with the kill-a-watt why not use VA and include the PS efficiency? IMO it is a worthy addition to overall system construction, anyone else for VA/TB?

When the power meter/billing/etc was devised long ago the vast majority of homes were almost purely resistive, billing based on watt made good sense. In today's world with just about everything having a switching type PS It would not surprise if they someday push for VA billing!
 
I am at 14W per TB when disks are idle
and 10W per TB when disks are spun down.

Its is a underclocked E5200 cpu with giga matx mobo, 3ware Raid card and 8x1TB WD green drives + one 500GB Green drive for OS. :)
 
as the wall plug power draw has little to do with how the drive are arranged

10 drives in a system should be able to goto sleep.......10 drives in a RAID6 will not, so there should be pretty big differences. My server with 7 1TB drives in, normally has 6 of them in sleep mode and thus using a lot less power as they are spun down.
 
This is a good idea

my 3.5 TB system currently pulls 170W

putting me at 48.5 W/TB

i will hope to upgrade that soon tho as currently there are loads of 400/500/750 GB drives there



something tells me im way over the norm on this however my system also runs VMs SQLserver and a couple of websites so maybe thats why!
 
I have about 5TB of data and use about 110w. 22w/TB.

Since I have 3 IDE drives and 8 SATA drives I could up my space to 13.5TB at the same power consumption - 8w/TB, but my data would still be about 5TB. Seems a great deal of expense to accomplish nothing.
 
Doesn't the CPU take far more power than the drives do?
 
Depends on how many drives you're talking about and what sort of CPU :)
 
Don't normally see MB wattage ratings. Take a common Core 2, rated at 65 watts/1.2v equals 54.16 amps!

A Barracuda 1 TB drive takes about 9 watts. A Core I7 system has a TDP of 135 watts, so you'd need 15 TB before the drives caught the CPU; or about 7 drives, for your Core2 example.

And that's counting only the CPU -- not the memory, or the graphics card, or any of the other things that will cause the PSU to pull power from the mains plug.
 
but the aim i suppose for some people is to produce a system that is as low power as possible with as much usable Space as possible (essentially no more than a NAS)

for a simple 10TB system throw 11HDDs at an intel atom and you would be looking at something like 100W divide that by ten to get 10W/TB which i suppose is a fair way of measureing a NAS' efficiency

it increases when you start neeeding a little extra grunt (Websites , VMs, different Raid Levels) that the W/TB ratio increases
i..e before you bring in any other equipment RAID 5 can be 50% less efficient @ W/TB than no raid levels
or a system that has a 65W CPU idleing @ 50% powerusage would be worse than a 135W CPU idling at 20% power usage

so the question is of the total system pull regardless of setup as it all adds to the basic NAS, it could make for some very interesting debate
 
10 drives in a system should be able to goto sleep.......10 drives in a RAID6 will not, so there should be pretty big differences. My server with 7 1TB drives in, normally has 6 of them in sleep mode and thus using a lot less power as they are spun down.

my WHS box runs the lights out add-in and goes to sleep when all the computers have turned off, then turns back on when a computer is detected.

so, the time averaged power draw is very low; though this doesn't affect the idle draw.
 
so the question is of the total system pull regardless of setup as it all adds to the basic NAS, it could make for some very interesting debate

Then just use RAID0 or JBOD to maximize the storage space, and the least demanding processor possible to minimize the power requirement. What's the debate?
 
10 drives in a system should be able to goto sleep.......10 drives in a RAID6 will not, so there should be pretty big differences. My server with 7 1TB drives in, normally has 6 of them in sleep mode and thus using a lot less power as they are spun down.

Why wouldn't the RAID6 setup be able to go to sleep or power down HDDs?:confused:
Sounds illogical IMO. Windows still se it as just one HDD...
 
Why wouldn't the RAID6 setup be able to go to sleep or power down HDDs?:confused:
Sounds illogical IMO. Windows still se it as just one HDD...

Maybe if there's absolutely no I/O going to the array, the very nature of RAID 6 and its striping will mean that all member disks will participate in whatever I/O there is for the array, so there's not that much chance for a RAID 6 to go idle.
 
Areca RAID controllers power down entire arrays after the set idle period. It works great for saving power.
 
Maybe if there's absolutely no I/O going to the array, the very nature of RAID 6 and its striping will mean that all member disks will participate in whatever I/O there is for the array, so there's not that much chance for a RAID 6 to go idle.

I understand. But if the array is not partitioned for OS usage, but rather only used as a D:\ drive I don't think it will have issues going to sleep.
Am I wrong on this?
I use my Raid5 this way and set the idle sleep time in windows. Works perfectly. How is Raid5 so much different than Raid6.
 
I understand. But if the array is not partitioned for OS usage, but rather only used as a D:\ drive I don't think it will have issues going to sleep.
Am I wrong on this?
I use my Raid5 this way and set the idle sleep time in windows. Works perfectly. How is Raid5 so much different than Raid6.

Hmm, I haven't really messed with Windows power save features on my home server, maybe I should do that as I don't have much use for the server during the day due to work and such, but I do use it in the evening. Might be able to save a few watts that way. :)

I've always tried to minimize the amount of spin ups/spin downs because I'm of the camp that spinning up/spinning down unnecessarily puts a lot of wear on the drive, though there's not a lot of scientific evidence of this at the moment,
 
both boxes from signature
13,5TB: ~120w idle resulting in 8,88W/TB or ~40w when all except sys drive are spindowned resulting in 2,96W/TB
19TB: ~135w idle 7,01W/TB or ~40w spindowned 2,1W/TB

on average, weekly watages are ~70-75Wh per box
that would be: 5,18W/TB and 3,94W/TB
 
seeing people in the >10TB thread posting their server power draw got me wondering...

for those with a power meter, what is the watts/TB value for your server?
only count storage included in the power measurement.
i have two external hard drives connected, but they're not counted in the below.

mine is

75w for 4.32TB

so

17.36 W/TB

You said you were in the UK, which uses 220v correct? This is also a factor because 75 watts at 220v would be the same as 150 watts for US users at ~110v.
 
i am in the UK, and the UK does indeed use 220v

though i believe the 75W draw on 220v would remain ~ 75W draw on 110v. all that would change is that the current on the 110v system required for the 75W draw would increase.

the higher current would lead to a slightly increased power draw due to increased ohmic heating in the psu. P=I^2 R

-----

from some of the systems posted it does seem that the mobo / cpu makes little difference. my atom based system does not seem much better in W/TB than other systems based on core 2 duo
 
Asposium, exactly. Btw it's 230V, not 220V anymore.

So 75W/230V = 0.326A, which in USA means 75W/110V = 0.68A. I know, it is a very simplistic approach, not taking into account the differences between AC and DC and other factors which can influence the result. But the main point is - 75W is still 75W.
 
Asposium, exactly. Btw it's 230V, not 220V anymore.

So 75W/230V = 0.326A, which in USA means 75W/110V = 0.68A. I know, it is a very simplistic approach, not taking into account the differences between AC and DC and other factors which can influence the result. But the main point is - 75W is still 75W.

whilst technically that is correct, PSUs are actually a little more efficient @ 230V so a system using 75W in the US would perhaps use 70W in the UK
 
Last edited:
A Barracuda 1 TB drive takes about 9 watts. A Core I7 system has a TDP of 135 watts, so you'd need 15 TB before the drives caught the CPU; or about 7 drives, for your Core2 example.

And that's counting only the CPU -- not the memory, or the graphics card, or any of the other things that will cause the PSU to pull power from the mains plug.

TDP isn't real draw. The Core i7 is a good example of this; when a core has no processing demand placed on it, it can be cut off entirely from power. So if you're sitting at idle all day (as many of these systems do) your power draw is much lower than if some process were taking up all the cycles it could.
 
Asposium
its 230v here
AMD 5050e cpu with cnq on gentoo
onboard graphics
Chieftec CFT-750-14CS psu

also ... 40w spindowned thingie should be noted, cause thats the way how those two babys spend most of their time :D

oh and Measurement Etech PM-300
 
Back
Top