Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.

alphaod

macrumors Core
Original poster
Feb 9, 2008
22,191
1,255
NYC
I recently decided to hook up a Kill-A-Watt to my Mac Pro; after metering for a while, I've calculated that my computer uses only 315W per hour when it's fully cranked; that's 100% CPU on all 8-cores, along with full CUDA cranking on the GTX 285.

And yes my CPUs are cranked 100%, 100% of the time.

Does this even sound right? I mean it's nice that's it's not using a lot power if the readings are right.
 
Apple said they're going greener... maybe they want their customers to be green too? Out of curiosity, how much power did you old machine draw?
 
I recently decided to hook up a Kill-A-Watt to my Mac Pro; after metering for a while, I've calculated that my computer uses only 315W per hour when it's fully cranked; that's 100% CPU on all 8-cores, along with full CUDA cranking on the GTX 285.

And yes my CPUs are cranked 100%, 100% of the time.

Does this even sound right? I mean it's nice that's it's not using a lot power if the readings are right.

If you leave out the bolded part out, I guess it should be right. I would've thought it might be a little higher, but I guess the new xeon's use less power and that the GPGPU portion doesn't actually use up the entire of the GPU's power like it does when it does gaming.

(a watt is a Joule/second, hence a watt/hour makes no sense)
 
Yeah... Goto bootcamp and load in something like 3DMark'09, pull out all the stops and run some intensive video tests, lets see how much wattage is being used then. I'm pretty sure the GPU and CPU will be fairly busy crunching stuff.
 
Apple said they're going greener... maybe they want their customers to be green too? Out of curiosity, how much power did you old machine draw?

The closest thing I used as a desktop was a Mac mini and that's not much of a desktop.

If you leave out the bolded part out, I guess it should be right. I would've thought it might be a little higher, but I guess the new xeon's use less power and that the GPGPU portion doesn't actually use up the entire of the GPU's power like it does when it does gaming.

(a watt is a Joule/second, hence a watt/hour makes no sense)

From my experience don't we we always measure electricity in kWh? Of course unless I've been getting everything wrong.

Yeah... Goto bootcamp and load in something like 3DMark'09, pull out all the stops and run some intensive video tests, lets see how much wattage is being used then. I'm pretty sure the GPU and CPU will be fairly busy crunching stuff.

Don't have Boot Camp ;)

Snow Leopard Server doesn't have that feature. And even when I install the client version, it gets to the installation part and then just does nothing, so no go for Boot Camp. :(

Of course I'm always up for suggestions on how to resolve that.
 
I recently decided to hook up a Kill-A-Watt to my Mac Pro; after metering for a while, I've calculated that my computer uses only 315W per hour when it's fully cranked; that's 100% CPU on all 8-cores, along with full CUDA cranking on the GTX 285.

And yes my CPUs are cranked 100%, 100% of the time.

Does this even sound right? I mean it's nice that's it's not using a lot power if the readings are right.

I fail to see why yours with a cranked up 8 cores and a GTX285 would only consume 315W when mine has an-almost similar spec MP minus a GTX285 consumed close to 300W on load.
 
From my experience don't we we always measure electricity in kWh? Of course unless I've been getting everything wrong.

How much electricity you use is measured in kW hours, not kW per hour. Watt = energy / time, watt x time = energy. The instant of energy you're using is measure in Watts (W).

The confusion might be from the wording you used in the OP but if you ran your MP at full bore, and it used an average of 315W over that hour, the reading on that Kill-A-Watt should say 0.315kWh (315W = 0.315kW, multiply by 1 hour of usage).

Also: I've never used a Kill-A-Watt thingy before, but does it actually record the total usage over time? In that case, you really ran your MP at max for a whole hour? And during that hour, was the usage consistent?
 
Also: I've never used a Kill-A-Watt thingy before, but does it actually record the total usage over time? In that case, you really ran your MP at max for a whole hour? And during that hour, was the usage consistent?

Yes it records over time. And it calculates the cost of electricity used.

I fold on my computer along with other things, so yes it's cranked up to almost max 24/7.
 
From my experience don't we we always measure electricity in kWh? Of course unless I've been getting everything wrong.

Yes we do, but that is how much was used, not the rate. You have to think of it this way: Watts = speed, KWh = distance.

Your computer is cruising along at 315W on the highway. After 1 hour it will have gone .315KWh. After 4 hours it will have gone 1.260 KWh.

KWh is a measurement of power used. That is how your power company charges you... based on how many KWh you consumed. 1KWh is equal to using 1000W for 1 hour, or 100W for 10 hour, or 10W for 100 hours.

And Watts per hour is not the same unit as KWh, or Wh.

(Watts/1000) * Hours = KWh
Watts/1000) * Hours / Hours = KW
KW * 1000 = Watts

Watts * hours = Wh
Wh/1000 = KWh

Watts/Hours = WTF? Actually, that would technically be a form of electrical acceleration. (Speed/time) m/s/s is acceleration. 10Km/h per hour means speeding up by 10Km/h every hour. So 10W per hour (10Joules/s/h) would mean using 10W more power every hour, after 10 hours you would be using 1000W and be blowing your power supply.

Can you believe I dropped out of engineering?

P.S I forget if 1 Watt equals 1 joule per second, or 1000 joules per second. So correct me if I am wrong. Been a while.
 
Yeah, KW/h and KW, and W are all different. But anyway, what you're getting is about right and the same my 2006 - (assuming you actually meant watts W)

According to Hardware Monitor I idle at around 235W average. It actually goes up and down like at heartbeat and about at the same pace, between 195W and 260W. At 100% usage of all CPUs it's about 345W and without the bobbing up and down. It still bobs but with only a 5W to 10W fluctuation instead of 65W to 70W like when idle.



EDIT: You can think of Watts as the water coming out of a hose, the bib-valve as ampere (current flow) and the water pressure as volts (electromotive force). Volts times Amps = Watts. kWh would be more like a bucket that catches the water. A measurement of how much water (power) came out of the hose in a 1 hour period.
 
Actually the proper acronym is kWh so whatever. :)

It's not *, /, |, +, or -. :D

Electricity_Counter.jpg


1 kWh =1000 x 3600 watt.secs = 1000 x 3600 joules = 3600 kilojoules. ;)

http://en.wikipedia.org/wiki/KWh I edited my post for you tho. :) Good catch!
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.