Maker Pro
Maker Pro

Typical mains power for mid-range PC?

J

Jon D

Jan 1, 1970
0
How much mains power does a modern systen unit need?

----

In more detail ... I am in the UK.

My existing PC (socket-A 462-pin cpu with 768 MB SD-RAM) uses about 180
Watts at 240 volts of which about 65 or 70 Watts is to power my CRT.
Printers and scanners would be extra.

Modern cpu's seem to be quite power hungry.

QUESTION: Approx how much mains power is likely to be needed for a
modern mid-range AMD-based PC? I don't know the existing AMD processors
but something average to middling is what I mean.

QUESTION: Would a sysem based on an Intel cpu need less power?
 
G

Gerard Bok

Jan 1, 1970
0
How much mains power does a modern systen unit need?

Trust me, they will require the same amount of power, wherever
you are. Only the pricetag will change.
My existing PC (socket-A 462-pin cpu with 768 MB SD-RAM) uses about 180
Watts at 240 volts of which about 65 or 70 Watts is to power my CRT.

65 to 70 is very decent for a CRT. If you want to save in this
field, look for a TFT.
Printers and scanners would be extra.

Modern cpu's seem to be quite power hungry.

Modern users are (CPU-)power hungry.
There is no free lunch, not even in processor-land.
You need to put energy in to get anything out.
QUESTION: Approx how much mains power is likely to be needed for a
modern mid-range AMD-based PC? I don't know the existing AMD processors
but something average to middling is what I mean.

There is no middling. Depends on how your PC is built and how you
use it. Some PCs switch down their CPU on low loads. But that
wouldn't do any good if you keep your PC busy.
QUESTION: Would a sysem based on an Intel cpu need less power?

No. Generally speaking, Intel is not your choise if you want to
conserve energy.
 
K

kony

Jan 1, 1970
0
How much mains power does a modern systen unit need?

----

In more detail ... I am in the UK.

My existing PC (socket-A 462-pin cpu with 768 MB SD-RAM) uses about 180
Watts at 240 volts of which about 65 or 70 Watts is to power my CRT.
Printers and scanners would be extra.

Modern cpu's seem to be quite power hungry.

QUESTION: Approx how much mains power is likely to be needed for a
modern mid-range AMD-based PC? I don't know the existing AMD processors
but something average to middling is what I mean.

It'll be close enough to what you already have, maybe 10W
more on average.

QUESTION: Would a sysem based on an Intel cpu need less power?

Depends on what you buy. Averaging Intel's entire line of
CPUs including Core 2 Duo, it'll be close enough to the AMD
alternative.

Intel's newest CPUs use less power but then their chipsets
use more, and then they want to focus on performance per
watt instead of watt per system. Nobody buys 1.3 systems so
it has to be performance per system if the important factor
is how much power a small number of systems uses.

In short, ignore power usage on normal (equivalently
equipped) desktops, if power usage is that important then
neither choice is suitable.
 
C

Cuzman

Jan 1, 1970
0
Jon D wrote:

" How much mains power does a modern systen unit need? "

http://extreme.outervision.com/psucalculator.jsp


" My existing PC (socket-A 462-pin cpu with 768 MB SD-RAM) uses about
180 Watts at 240 volts of which about 65 or 70 Watts is to power my CRT.
Printers and scanners would be extra. "

The CRT, printer and scanner should have their own power units
independent of the PC's PSU.

Make a list of everything that isn't powered by the PC's PSU. Find the
specs on the manufacturer's websites and add them to the total wattage
you need for the PC PSU (see aforementioned link to calculator). Work
out the total wattage, divide by 1000, then times by 0.0633 (averagely
high cost of electricity) . That will tell you the approximate maximum
costs to run your PC each hour (in £s).

By that logic, a total of 470W would cost less than 3 pence per hour (
(470 / 1000) * 0.0633 = 0.029751)

Cost of electricity: http://www.ukenergy.co.uk/pages/calculation.html

http://www.bbc.co.uk/schools/gcsebi...ity_and_magnetism/mains_electricityrev7.shtml

Bear in mind that, in reality, the cost will be considerably less than
that, as much of the setup would not be running constantly and under duress.


" Modern cpu's seem to be quite power hungry. "

Yes and no. It depends how you look at the argument. They have more
capabilities than previous generations of CPUs, but big steps have been
made in nanometre architecture. From the dawn of the Athlon XP the
process has gone from 180nm to 130nm to 90nm and now towards 65nm.

If you were to re-encode the same 2 hour video on both an Athlon XP and
an Athlon 64 X2 you would find that the Athlon 64 X2 would be using more
power, but it would also do the task considerably faster. If a system
uses twice the power, but does the task in half the time, what's the
difference? You would only end up using more power by using your PC
more (which often becomes the case)

Modern systems can use considerably more power for games because the
graphics cards and CPUs can output more frames per second (FPS). If
everyone set their systems to limit the FPS they would find their power
usage to be lower. However, most people don't do this, allowing their
systems to use as much power and capability as their components will
give them.

DDR2 actually uses less power than DDR(1), but yet operates at higher
bandwidths.
http://www.corsairmemory.com/corsair/products/specs/ddr2_faq.pdf

Another area of power consumption has been the rise in spindle speed for
hard drives. The rise from 5400rpm to 7200rpm has been a large factor
in power increases, and many people even have 10,000rpm hard drives in
their desktop systems now. However, as with other components, steps
have been made to limit the power used by read/write operations.

One other big factor is the move towards Active PFC (Power Factor
Correction) in PSUs. http://www.endpcnoise.com/cgi-bin/e/pfc.html


" Approx how much mains power is likely to be needed for a modern
mid-range AMD-based PC? I don't know the existing AMD processors but
something average to middling is what I mean. "

Mid-range means different things to different people. CPU speeds often
dictate how "modern" a PC is, but everyone has different requirements in
their RAM requirements, graphics card usage/capabilities and the
size/number of hard drives they wish to use. Everyone also has
different amounts of add-on equipment and other stuff.


" Would a sysem based on an Intel cpu need less power? "

Not necessarily. Intel came under some criticism because their 90nm
desktop Prescott CPUs used more power than the equivalent AMD CPUs which
were still at 130nm. However, Intel have had more success recently with
65nm, and it's not like all Intel CPUs use more power than all AMD CPUs.


If you're that worried about power consumption you could opt for a 35W
Athlon 64 X2 3800+.
http://www.amdzone.com/modules.php?op=modload&name=Sections&file=index&req=viewarticle&artid=257

You'll find that article quite interesting where it states the total
power output under load as being 82W, but don't think for a minute that
you only need use a 100W PSU. In the review above they use a 500W
CoolerMaster iGreen.
http://www.coolermaster-europe.com/...guage_s=2&url_place=product_list&p_class=1114

With the following components you could have a gaming PC that will
certainly last for a couple of years (or even longer), and which uses a
little less power than the comparable systems:

- AMD Athlon 64 X2 3800+ 35W AM2 ADD3800IAA5CU
(see previous review)
- Asus M2N32-SLI Deluxe motherboard
(just because it's currently the best AM2 motherboard)
- Corsair XMS2 2x1GB DDR2-800 TWIN2X2048-6400
(1.9v operating voltage is low for DDR2-800)
- Nvidia GeForce 7900GT 256MB
(hits a sweet spot with power/performance
http://www.vr-zone.com/?i=3335&s=8 )
- Seagate Barracuda 7200.10 320GB ST3320620AS
(lots of storage, sweet spot in price, reliable, 5 year warranty)
- CoolerMaster iGreen 430W RS-430-ASAA
(see aforementioned link to CoolerMaster website)

....and just as important, add an LCD TFT monitor.
 
D

david schwerbel

Jan 1, 1970
0
|> How much mains power does a modern systen unit need?
.....
| Dorothy Bradbury

very nice post, good work
 
R

Rod Speed

Jan 1, 1970
0
For HD I stated...
o 7200rpm 3.5" -- 10W at idle,

And that is overstating it.
some more, 20-25W maximum

And that is too.
o 10000rpm 3.5" -- assume 30W re cooling needs

Bad assumption.
7200rpm 3.5" manuf'r specs...
o Current hard drives
---- smallest capacity -- 7.50W idle, 12.5W active
---- largest capacity -- 9.30W idle, 13.3W active

Those are overstated on the idle.
---- variable quality -- 8.10W idle, 13.6W active

God knows what 'variable quality' is about.
o Older HDs can have somewhat higher figures
o Overall I would still take 10W at idle, 20W max if older HD

Pity that none of the other numbers are with older hardware.
10000rpm 3.5" manuf'r specs...
o Current hard drives
---- high performance actuator -- idle 7.9W, 8.4W active
---- very high performance actuator -- idle 12.2W, 17.4W active

Nothing like your previous 30W
o Trends
---- 1) higher performance actuators to justify price premium (watts)
---- 2) silicon disks forcing higher cache on electromag disks (watts)
Waffle.

o Overall I would still assume 30W re cooling needs

More fool you.
In terms of energy usage the confounding factor in PCs is not the
HD/RAM/Optical or even CPU (P4 Prescott 115W v Conroe 50W).

That last is a good example of the problem with your
original and you didnt even mention the AMD cpus.
It is the actual GPU fitted - integrated may be 10W,

Again, nothing like your original.
whereas a high end card can draw well over 100W & idle not far off that figure.
Saving 50W by choosing a very pricey m/b & chip does not gain
much if you also u/g the graphics card to one drawing 50W more.

Having fun thrashing that straw man ?
Likewise a CD-ROM may draw only a few watts, but a high end
DVD-R/RAM/CDR writer multi-drive can draw 18-25W at peak.

Not very often tho.
Yes CPUs can draw a lot of power - but so do other devices.

Duh.
 
R

Rod Speed

Jan 1, 1970
0
John Jordan said:
Rod Speed wrote
Easily tested.

Not really, most dont have any way of measuring the power their PC takes.
Put PC on wattmeter, test idle with one drive. Add second 7200rpm IDE drive, test idle
and folder-
copy (within second drive), subtract to get difference.
With a 120GB Maxtor plus-9 (from the "probably dying" pile), I get:
idle: 10W

Thats nothing like typical for 7200 rpm drives.
load: 24W

And neither is that.
Note that these are mains-side wattages, so they include PSU inefficiency.

So arent relevant to Dorothy's numbers.
Idle is spinning-idle, as hardly anyone spins down non-laptop drives.

Unlikely given that most modern OSs default to spinning them down.
 
R

Rod Speed

Jan 1, 1970
0
John Jordan said:
Rod Speed wrote
It's an extremely common drive (at least in the UK),

Most current drives dont idle at 10W.
and far from the most thirsty (I have an early 7200rpm Seagate somewhere that uses
around twice the idle power).

Dont believe it, list the model number.
My 120GB Seagate has similar idle consumption to the Maxtor but somewhat less load.
Dorothy didn't explicitly state whether the wattages in the original post were mains or
PSU-side,

That was obviously implied. And should have been
obvious from the Prescott cpu number cited.
and you never asked. The typical difference would be around 30%, in case you care.

I dont, because that wasnt what her numbers were about.
 
C

Cuzman

Jan 1, 1970
0
Rod Speed wrote:

" Most current drives dont idle at 10W.... ....Dont believe it, list
the model number. "



Select the drop-down menu. Choose each of the bottom four tests and
sort. The second resulting table lists the numbers clearer.

http://www.storagereview.com/comparison.html

That should put the cat among the pigeons.
 
R

Rod Speed

Jan 1, 1970
0
Cuzman said:
Rod Speed wrote
Select the drop-down menu. Choose each of the bottom four tests and sort. The second
resulting table lists the numbers clearer.

That should put the cat among the pigeons.

Nope, even someone as stupid as you should be able to work out
that the idle power with commonly used drives is nothing like 10W.

In spades with Dorothy's stupid 20-25W max.
 
W

Wes Newell

Jan 1, 1970
0
that the idle power with commonly used drives is nothing like 10W.

In spades with Dorothy's stupid 20-25W max.

If you check the specs from the manufacturers websites, you''l find that
most current drives do in fact have idle power dissipation around 10W.

Mainstream drives.

150MB/s sata drives;

Power Dissipation
Read/Write 9.00 Watts
Idle 8.75 Watts
Standby 1.20 Watts
Sleep 1.20 Watts

300MB/s sata drives;

Power Dissipation
Read/Write 9.50 Watts
Idle 8.75 Watts
Standby 1.60 Watts
Sleep 1.50 Watts

100MB/s eide drives.

Power Dissipation
Read/Write 8.60 Watts
Idle 8.10 Watts
Standby 1.10 Watts
Sleep 1.00 Watts

http://www.wdc.com/en/products/products.asp?driveid=41&language=en

And from Seagate.

Typical Current (12VDC +/- 5%) 0.433 amps
Startup Current (12VDC +/- 5%) 2.8 amps
Typical Current (5VDC +/- 5%) 0.401 amps
Startup Current (5VDC +/- 5%) 0.841 amps
Power
Seek Power (typ) 12.4 watts
Read/Write Power (typ) 12.8 watts
Idle Power (typ) 7.2 watts


It seems their startup power would be about 38W (2.8*12)+(.8*5). This is
the standard Baracuda drive family. So it seems like Dot may be a little
low with only 20-25W.

http://www.seagate.com/cda/products/discsales/enterprise/tech/1,1084,701,00.html
 
C

Cuzman

Jan 1, 1970
0
Rod Speed wrote:

" Nope, even someone as stupid as you should be able to work out that
the idle power with commonly used drives is nothing like 10W. In spades
with Dorothy's stupid 20-25W max. "



I never agreed or disagreed with either of you. I just pointed out that
it would be interesting to throw those benchmark results into the
argument. They are done by a team of people likely to have more
expertise on the matter than anyone in this argument.

Here are some more benchmark results out there which you can use:

http://www.digit-life.com/articles2/storage/hddpower.html
 
R

Rod Speed

Jan 1, 1970
0
Wes Newell said:
Rod Speed wrote
If you check the specs from the manufacturers websites, you''l find that
most current drives do in fact have idle power dissipation around 10W.

Says he ignoring the storagereview data.

The reason for that is that the datasheet numbers are worst case numbers.
Mainstream drives.

150MB/s sata drives;

Power Dissipation
Read/Write 9.00 Watts
Idle 8.75 Watts
Standby 1.20 Watts
Sleep 1.20 Watts

300MB/s sata drives;

Power Dissipation
Read/Write 9.50 Watts
Idle 8.75 Watts
Standby 1.60 Watts
Sleep 1.50 Watts

100MB/s eide drives.

Power Dissipation
Read/Write 8.60 Watts
Idle 8.10 Watts
Standby 1.10 Watts
Sleep 1.00 Watts

http://www.wdc.com/en/products/products.asp?driveid=41&language=en

And from Seagate.

Typical Current (12VDC +/- 5%) 0.433 amps
Startup Current (12VDC +/- 5%) 2.8 amps
Typical Current (5VDC +/- 5%) 0.401 amps
Startup Current (5VDC +/- 5%) 0.841 amps
Power
Seek Power (typ) 12.4 watts
Read/Write Power (typ) 12.8 watts
Idle Power (typ) 7.2 watts

None of those are 10W or the 20-25W max in spades.
It seems their startup power would be about 38W (2.8*12)+(.8*5).
This is the standard Baracuda drive family. So it seems like Dot
may be a little low with only 20-25W.

http://www.storagereview.com/comparison.html
 
R

Rod Speed

Jan 1, 1970
0
Cuzman said:
Rod Speed wrote
I never agreed or disagreed with either of you.

You did make that silly claim about the cat and pigeons.
I just pointed out that it would be interesting to throw those benchmark results into
the argument. They are done by a team of people likely to have more expertise on the
matter than anyone in this argument.

It aint about expertise, just bothering to measure it.
Here are some more benchmark results out there which you can use:

Which dumps on Dorothy's numbers too, like I said.

And suffers from the problem that datasheets always
have, the datasheet numbers are worst case numbers.
 
R

Rod Speed

Jan 1, 1970
0
o Go to any of the HD maker websites
o Download the manufacturers specification
o Read it.

Datasheet numbers are worst case numbers and
STILL most arent 10W with current desktop drives.

What makes a lot more sense is
http://www.storagereview.com/comparison.html
which shows nothing like what you claimed.

In spades with the max number.
Stick an ammeter on a current Barracuda & older Barracuda.

There might just be a few more drives around than those.
Do not use the 2A fused range during startup obviously.
PCs have a range of wattage draw re idle/loaded...
o Base end office Cel2.0, integrated graphics, 1 HD, 1 optical
o idle -- 85W with 17" TFT, 95W with 19" TFT
o load -- 120W for 17" TFT, 130W with 19" TFT

Irrelevant to that bit about DVD-R/RAM/CDR writers.
You really are a troll.

You really are a pathetic excuse for a bullshit artist.
HD spec of 8-17W can not be applied to cooling needs...

Depends entirely on the case design.
o 8-17W is not distributed evenly over the whole drive
---- a HD is Actuator, Spindle Assembly, PCB Components

Irrelevant to your silly claim about 30W
o 8-17W may be focused in specific components
---- with specific maximum operating temps
---- requiring specific linear feet per min of airflow

Irrelevant to your silly claim about 30W
o Providing cooling for 8-17W ignores local needs
---- thermal shadowing by other components
---- thermal shadowing by case design & HD positioning

Irrelevant to your silly claim about 30W
Stick 2x 10,000rpm HD against one another in a drive cage.

Only the stupid do something as stupid.
Provide sufficient cooling for 2x 8-17W drives.

Or dont have them adjacent, stupid.
I will provide the data recovery firm name to use...

I'm not stupid enough to need them.
o Motor-IC overheats from lack of linear-feet-per-min airflow
o Motor-IC dumps heat into surroundings, motor/spindle heat
o Motor-spindle dumps heat into FDB accelerating aging
o FDB aging causes higher current draw (startup & continuous)
---- S.M.A.R.T. data measures startup Ima &/or startup Tsecs
o Higher current draw heats motor/spindle & Motor-IC
---- Reinforcing System Loop of increasing thermal dissipation

Only in that terminally stupid config.
Same thing happens more slowly with HDs anyway - hence the
older a HD the higher the startup current & higher op current.
Add another HD because the old one is too small, fit them too
close & you can accelerate the demise of the old one by temps.

Only the stupid do things that stupidly.
I would assume 30W re cooling needs for 10k-rpm HDs.

You can assume anything you like. Anyone
with a clue doesnt use such a stupid config.
You can assume what you wish.

I dont assume, I measure, thanks.

Its a very simple sentence, even you should be able to manage it.
Yes but you have to assume it at SOME point re 1) cooling 2) PSU.
Nope.


Ah now I see. Intel tend to be worst case from a thermal perspective.

Wrong, as always.
 
R

Rod Speed

Jan 1, 1970
0
Fuckwit.

John Doe said:
Troll


Path:
newssvr25.news.prodigy.net!newsdbm05.news.prodigy.com!newsdst01.news.prodigy.net!prodigy.com!newscon06.news.prodigy.com!prodigy.net!fu-berlin.de!uni-berlin.de!individual.net!not-for-mail
From: "Rod Speed" <rod.speed.aaa gmail.com>
Newsgroups:
alt.comp.hardware.overclocking.amd,uk.comp.homebuilt,alt.comp.hardware,alt.engineering.electrical
Subject: Re: Typical mains power for mid-range PC?
Date: Sat, 22 Jul 2006 19:46:27 +1000
Lines: 31
Message-ID: <4ieaflF28o6pU1 individual.net>
References: <Xns980762F27AD4017E53A 127.0.0.1> <Zg4wg.13298$u%3.1572
newsfe1-gui.ntli.net> <e9r9sm$jjk$1 cache.uni-koblenz.de>
<4icu0tF39bmfU1 individual.net> <STbwg.38082$1g.25247
newsfe1-win.ntli.net> <4id5rfF37s96U1 individual.net>
<Qoewg.51400$ST2.40253 newsfe5-win.ntli.net> <4idh8hF370ceU1
individual.net> <4idkd6F3b8t2U1 individual.net> <4idlkjF3bg74U1
individual.net> <4idobhF3bjhgU1 individual.net> <4idqhcF3c55oU1
individual.net> <e9spi6$2bv$1 news.freedom2surf.net> X-Trace:
individual.net kTaYagFJ9PlFFtgBvVDVpwyGyxfBcSHgy0atJc9sqr24c7sg8=
X-Priority: 3
X-MSMail-Priority: Normal
X-Newsreader: Microsoft Outlook Express 6.00.2900.2670
X-MimeOLE: Produced By Microsoft MimeOLE V6.00.2900.2670
X-RFC2646: Format=Flowed; Response
Xref: prodigy.net alt.comp.hardware.overclocking.amd:269655
alt.comp.hardware:314825 alt.engineering.electrical:179563




Nope, even someone as stupid as you should be able to work out
that the idle power with commonly used drives is nothing like 10W.

In spades with Dorothy's stupid 20-25W max.
 
Top