Anyone here got any experience with a Kill-A-Watt Device? [Archive]

cmcquistion

10-08-03, 11:25 PM

dustybyrd

10-08-03, 11:33 PM

Has anyone here used one?

It is a device you plug into the wall, then plug your appliance/computer/whatever into it. You run the appliance/computer/whatever and the Kill-A-Watt device has a readout that can tell you Volts, Amps, Watts, VA (voltage * current value, which includes harmonics), Hz, PF, and KWH.

I just got one. I've been messing around with it a bit. I was hoping to find some others that have used it and can help me with some questions and give some comparison figures.

larva, please tell me you've used these and can help me....

i will be hoping you will tell us what you find on power consumption....

particularly on your dual AMD systems (under full load and no load...and o/c'd)

please, please...

i have read about someone using this in the 2cpu.com forums...in the cases and power supply section...i will look back into it

cmcquistion

10-08-03, 11:36 PM

Originally posted by dustybyrd
i will be hoping you will tell us what you find on power consumption....

particularly on your dual AMD systems (under full load and no load...and o/c'd)

please, please...

i have read about someone using this in the 2cpu.com forums...in the cases and power supply section...i will look back into it

You can find my experience, thus far, in this evil forum (http://forums.silentpcreview.com/viewtopic.php?t=7553). The first system is a simple single-CPU Palomino XP1600 machine at 10.5 * 166 @ 1.75 Vcore. The second, high wattage system I am talking about in that post is my AMD dually @ 15 * 150 @ 1.675 Vcore (which isn't totally stable... I need to bump the Vcore up to 1.725, since the board doesn't offer 1.70). It uses a lot of juice, but I'm not sure, yet, what all the numbers equate to.

*EDIT* It was poor taste of me to refer to the SPCR forum as "the evil forum". I apologize. I should have referred to it as the SilentPCReview forum, which is unmoderated and does not seem to have any enforced rules on conduct.

*EDIT 2* Some of the wattage numbers above, are out of date and are adjusted in posts, below, due to the system not being fully stable at the previously tested Vcore, and the addition of another stress test (3DMark2003).

larva

10-08-03, 11:42 PM

No, I haven't used one, but I have measured many things with many detection devices. What sort of questions do you have? (I can tell you right now the numbers will be shockingly low)

cmcquistion

10-08-03, 11:45 PM

Originally posted by larva
No, I haven't used one, but I have measured many things with many detection devices. What sort of questions do you have? (I can tell you right now the numbers will be shockingly low)

I'm so glad you dropped in. I trust your knowledge of PSU's more than anyone else.

Here are my questions, asked in the evil forum (http://forums.silentpcreview.com/viewtopic.php?p=46991#46991). I don't know, yet, whether my questions are even on the right track.

I have another system that is using 294 watts at full load, according to the Kill-A-Watt Meter, and I know that the PSU's efficiency is 68% (minimum, according to the manufacturer, at max power.) Does this means that the components in the system are really drawing 200 watts (or more, depending of true efficiency), and the PSU is producing 94 watts of heat, which is the loss in the efficiency of the AC/DC conversion?

Let's assume I wanted to replace this PSU with a different one, or build an identical system in all respects, except for the PSU. Assuming the PSU for the other system is just as efficient as this PSU, does the PSU for the other system need to be rated for at least 200 watts, based on its total output of 3.3V, 5V, and 12V lines (or just the 3.3V and 5V)? Or, does it need to be rated for at least 294 watts, based on its total output of 3.3V, 5V, and 12V lines (or just the 3.3V and 5V)?

Am I asking the right questions?

This is an overclocked AMD dually, by the way.

*EDIT* It was poor taste of me to refer to the SPCR forum as "the evil forum". I apologize. I should have referred to it as the SilentPCReview forum, which is unmoderated and does not seem to have any enforced rules on conduct.

*EDIT 2* Some of the wattage numbers above, are out of date and are adjusted in posts, below, due to the system not being fully stable at the previously tested Vcore, and the addition of another stress test (3DMark2003).

dustybyrd

10-08-03, 11:58 PM

(I can tell you right now the numbers will be shockingly low) [/B]

if i remember correctly, a dual xeon 2.4ghz system uses about 175-250 watts from no load to full load conditions...from 2cpu.com forums...

but i do believe this was real watts...thus at 65% efficiency that would mean a 385 watt power supply minimum would be required to drive that system at full load...

and i am not sure whether there were many drives running or high end video cards at the same time...

cmcquistion

10-09-03, 12:06 AM

I have some updated numbers:

Since my dual @ 1.675 Vcore wasn't totally stable, I had to bump the Vcore up to 1.75 (the K7D doesn't have a Vcore option of 1.70 for some reason.)

My full load wattage, now, is 320 watts. The efficiency of the PSU is 68% (minimum at 520 watts). Since I am operating below the maximum wattage, I assume my efficiency is probably higher than that.

*EDIT* I had to bump up the Vcore to 1.75. The full load wattage, here, is when running Prime95 and BurnInTest, but not 3DMark2003.

dustybyrd

10-09-03, 12:12 AM

I have some updated numbers:

Since my dual @ 1.675 Vcore wasn't totally stable, I had to bump the Vcore up to 1.725 (the K7D doesn't have a Vcore option of 1.70 for some reason.)

My full load wattage, now, is 312 watts. The efficiency of the PSU is 68% (minimum at 520 watts). Since I am operating below the maximum wattage, I assume my efficiency is probably higher than that.

so if i understand correctly, that would mean that the system is requiring about 210 watts at full load?

what if you tried to burn (or run) a cd, use all available hard drives, and high end graphics at the same time as 2x prime 95 instances?

what about using passmark's burnin tester (that stresses all components at the same time)?

cmcquistion

10-09-03, 12:14 AM

Originally posted by dustybyrd

so if i understand correctly, that would mean that the system is requiring about 210 watts at full load?

what if you tried to burn (or run) a cd, use all available hard drives, and high end graphics at the same time as 2x prime 95 instances?

what about using passmark's burnin tester (that stresses all components at the same time)?

yeah, it is using about 212 watts (or higher, depending on the PSU's true efficiency).

Strange you should mention Passmarks BurnInTest. That is what I'm using, along with two instances of Prime95. I have CD-ROMs in both drives and I'm testing both hard drives, too.

I suppose I could try running 3DMark, too. (I love having a dually. Imagine trying to run all these programs on a single-CPU machine.)

dustybyrd

10-09-03, 12:19 AM

yeah, it is using about 212 watts (or higher, depending on the PSU's true efficiency).

Strange you should mention Passmarks BurnInTest. That is what I'm using, along with two instances of Prime95. I have CD-ROMs in both drives and I'm testing both hard drives, too.

I suppose I could try running 3DMark, too. (I love having a dually. Imagine trying to run all these programs on a single-CPU machine.) [/B]

please run 3dmark too :)

in addition to 2xprime95 and passmark, wow that's a lot at once...i couldn't imagine ever drawing more power than that...

also, you said you were using 1.675 to 1.725 volts...what cpu speeds?

what about full system specs please?

larva

10-09-03, 12:43 AM

Originally posted by cmcquistion
I have another system that is using 294 watts at full load, according to the Kill-A-Watt Meter, and I know that the PSU's efficiency is 68% (minimum, according to the manufacturer, at max power.) Does this means that the components in the system are really drawing 200 watts (or more, depending of true efficiency), and the PSU is producing 94 watts of heat, which is the loss in the efficiency of the AC/DC conversion?
When we talk of efficiency in this application (and in nearly all applications), we describe the relationship between input and output. When we say a supply is 70% efficient, we mean the output power is 70% that of the input. So yes, a supply operating at 70% efficiency and drawing an input of 294W is outputing 294(.7)W ~ 206W. Told you the numbers are low.

Originally posted by cmcquistion
Let's assume I wanted to replace this PSU with a different one, or build an identical system in all respects, except for the PSU. Assuming the PSU for the other system is just as efficient as this PSU, does the PSU for the other system need to be rated for at least 200 watts, based on its total output of 3.3V, 5V, and 12V lines (or just the 3.3V and 5V)? Or, does it need to be rated for at least 294 watts, based on its total output of 3.3V, 5V, and 12V lines (or just the 3.3V and 5V)?
Don't confuse the rating of a power supply with its output. They are not the same thing. Just because a power supply might be rated at 200W does not mean it wouldn't melt under what you suspect is an actual 200W load. You can't power a PC with a spec sheet, and the reality is a lot dirtier and unpredictable than those searching for easy answers wish to believe.

In general, numbers simply don't matter. It isn't until you experimentally determine (measure) power consumption that you gain any insight into power supply capacity. Ratings are the weakest form of psuedo-knowledge, and cannot be compared to actual measurements as they are not bound by the same unit definitions and testing standards.

Also understand that power supply efficiency varies a great deal between supplies of different types and manufacturers. 70% is very efficient for a switching power supply, and it is only in recent years that devices of this nature have reached this high an efficiency level. In years past 50% was a good rule of thumb for efficiency, but 70% is realistic for a quality supply benefiting from the wizard-like optimization found in-house at the various Taiwanesian power supply makers. I don't think it is realistic to expect efficiency to rise much if any beyond 75% with the current state of the art.

cmcquistion

10-09-03, 12:50 AM

I just ran a round of 3DMark2003, while running BurnInTest and two instances of Prime95. (15 * 150 FSB @ 1.75 Vcore)

I had spikes up to 357 watts (37 watts higher than it runs without running 3DMark).

Now, my question is this. If trying to determine how many watts a PSU needs to drive this system, should I take efficiency into account, as MikeC suggests, or not?

*EDIT* larva answered this, while I was typing this response.

Next, when comparing wattages of PSU's, should I use the Nominal Wattage (3.3V*Amps + 5V*Amps) or the Max Wattage (3.3V * Amps + 5V * Amps + 12V * Amps)?

*EDIT* I think larva may have answered this, too, while I was typing this response. If I understood correctly, the basic rule of thumb is... you can't tell because the PSU manufacturer can't be trusted to provide good relative data???

dustybyrd, my full system specs are this:

MSI K7D Master
2 Barton XP2500's, overclocked to 15 * 150 FSB @ 1.725 Vcore
dual SLK-800 heatsinks with Panaflo H1A fans
2 - 512 MB sticks of Registered ECC PC2700
Vantec Stealth 520 Watt PSU (not recommended, too loud) (I modded mine)
floppy drive
Lite-On DVD+RW
LG Electronics DVD/CD+RW
Seagate 15K.3 SCSI hard drive
Seagate Barracuda V ATA hard drive
LSI U160 SCSI card
Gainward GF4 Ti4600 video card
Audigy 2 soundcard

all housed in a Compucase LX-6A19,
with one exhaust fan-> 120mm*38mm Panaflo FBK-12G12LH (OEM Panaflo, pretty quiet),
and one intake fan-> PCPower& Cooling Silencer (the quietest 80mm fan I have ever used)
(this 80mm fan is elastic-suspended in a 120mm hole. Really it mainly operates as a hard drive cooler. The negative case pressure caused by the 120mm exhaust fan is not a problem. It sucks air right through the 120mm front intake hole, to compensate for the negative case pressure it causes. The front bezel has a dust filter, which is nice for keeping the system clean.)

Every fan in the system (2 heatsink fans, 1 PSU fan (another Silencer), one intake fan, and one exhaust fan) are on a rheobus and fan speeds are lowered to tolerable noise levels when the system is run at it's normal "quiet-mode", which is 15 * 133 FSB @ 1.40 Vcore. When overclocked to the max, the fans are run full speed.

*EDIT* When run in quiet-mode (15 * 133 @ 1.40 Vcore) the system draws a max of 233 watts.

*EDIT 2* My previous numbers were changed for the overclocked system, because it was not fully stable at 1.725 Vcore and had to be bumped up to 1.75.

cmcquistion

10-09-03, 01:10 AM

dustybyrd

10-09-03, 01:14 AM

I just ran a round of 3DMark2003, while running BurnInTest and two instances of Prime95. (15 * 150 FSB @ 1.725 Vcore)

I had spikes up to 343 watts (31 watts higher than it runs without running 3DMark).

Now, my question is this. If trying to determine how many watts a PSU needs to drive this system, should I take efficiency into account, as MikeC suggests, or not?

*EDIT* larva answered this, while I was typing this response.

Next, when comparing wattages of PSU's, should I use the Nominal Wattage (3.3V*Amps + 5V*Amps) or the Max Wattage (3.3V * Amps + 5V * Amps + 12V * Amps)?

*EDIT* I think larva may have answered this, too, while I was typing this response. If I understood correctly, the basic rule of thumb is... you can't tell because the PSU manufacturer can't be trusted to provide good relative data???

dustybyrd, my full system specs are this:

MSI K7D Master
2 Barton XP2500's, overclocked to 15 * 150 FSB @ 1.725 Vcore
dual SLK-800 heatsinks with Panaflo H1A fans
2 - 512 MB sticks of Registered ECC PC2700
Vantec Stealth 520 Watt PSU (not recommended, too loud) (I modded mine)
floppy drive
Lite-On DVD+RW
LG Electronics DVD/CD+RW
Seagate 15K.3 SCSI hard drive
Seagate Barracuda V ATA hard drive
LSI U160 SCSI card
Gainward GF4 Ti4600 video card
Audigy 2 soundcard

so i guess that your system power requirement would be very similar to mine...

which to me means you need a psu capable of continuously outputing about 350 watts....(because i would think that although the system is only needing ~235 watts....the psu can not supply that without some loss of energy as heat...thus 350watts)....

but then you need some safety headroom, i would think...because a 350 watt power supply, if it could power that system (which i doubt it could based on my own experience with the 350 watt fortron and my duallie) would be maxed out all the time...that couldn't be good on the psu...or voltage stabilities...

thus i would think you would need at least a good 400+watt psu to ensure safe use under super full load...

also, when you ran all that stuff at once, what settings did you have the passmark burnin test set on? 50% on each? (standard setting) or 90-100% on each component?

also, did you stress both hard drives and optical drives at the same time with that stress test?

cmcquistion

10-09-03, 01:16 AM

Originally posted by dustybyrd
also, when you ran all that stuff at once, what settings did you have the passmark burnin test set on? 50% on each? (standard setting) or 90-100% on each component?

also, did you stress both hard drives and optical drives at the same time with that stress test?

Yeah, I had all the Passmark BurnInTest settings at their default 50%, since I knew I would be running all the tests, and some other apps (two instances of Prime95 and 3DMark2003.)

I did stress both hard drives and both optical drives (and the floppy drive), as well as the memory, CPU, 2D video, 3D video, etc... (basically, everything but the printer, since I don't have one)

dustybyrd

10-09-03, 01:18 AM

According to my CPU Power calculator, my CPU's, when run at 2250 MHz @ 1.725 Vcore are pulling 91.6 watts a piece. This comes up to 183.2 watts, together. If my system is only pulling 343 watts, max, at 68% efficiency, then that equates to 233 watts. This would mean that after you subtract the wattage from the CPU's, the rest of the system is only using 50 watts.

This doesn't sit quite right, especially considering that I had to run 3DMark to get the wattage up that high. Previous to 3DMark, the system was only pulling 212 watts (312 watts * 68%), but the CPU's were at full load, pulling 182.6 of those watts, therefore the system was only using 29.5 watts!

Something seems fishy about these numbers.

what is your cpu power calculator?

also, if you go to amd's web site on the power requirement of the cpus...they list the max and nominal....so maybe that 91 watts/cpu is the max it would ever hit at that speed and voltage...but nominally it uses 70-75 watts/cpu...

which would leave 70-80 watts for the rest of the system..

cmcquistion

10-09-03, 01:21 AM

Originally posted by dustybyrd
What is you cpu power calculator?

also, if you go to amd's web site on the power requirement of the cpus...they list the max and nominal....so maybe that 91 watts/cpu is the max it would ever hit at that speed and voltage...but nominally it uses 70-75 watts/cpu...

which would leave 70-80 watts for the rest of the system..

I'm using the "CPU Power by Kostik" available here (http://www.silentpcreview.com/modules.php?op=modload&name=Web_Links&file=index&req=viewlink&cid=19).

I listed 91.6 watts because I am overclocking my CPU's, both in speed and Vcore. This CPU Power Calculator program lets you plug in your CPU model, then the speed and Vcore. It computes how many watts your CPU will use at that speed and Vcore.

I'm assuming that I am using max power, not nominal power, since the CPU's are both pegged to 100%, while I'm running BurnInTest, 2 Prime95's, and 3DMark2003.

larva

10-09-03, 01:28 AM

Originally posted by cmcquistion
According to my CPU Power calculator, my CPU's, when run at 2250 MHz @ 1.725 Vcore are pulling 91.6 watts a piece. This comes up to 183.2 watts, together. If my system is only pulling 343 watts, max, at 68% efficiency, then that equates to 233 watts. This would mean that after you subtract the wattage from the CPU's, the rest of the system is only using 50 watts.

This doesn't sit quite right, especially considering that I had to run 3DMark to get the wattage up that high. Previous to 3DMark, the system was only pulling 212 watts (312 watts * 68%), but the CPU's were at full load, pulling 182.6 of those watts, therefore the system was only using 29.5 watts!

Something seems fishy about these numbers.

When you calculate your cpus' power consumption, don't you start with a rating? Ratings are not intended to reflect the actual operatiing conditions in this case, rather define the outer limits. Actual cpu power consumption must be measured rather than estimated, as the error contained in the estimation is significant enough to bring the whole thing down to guessing.

Just like with the power supply itself, if you put a oscilloscope on the board and measure the consumption of the cpus you will find it lies at a much lower level than common wives tales, estimates and ratings dictate.

cmcquistion

10-09-03, 01:31 AM

Originally posted by larva
When you calculate your cpus' power consumption, don't you start with a rating? Ratings are not intended to reflect the actual operatiing conditions in this case, rather define the outer limits. Actual cpu power consumption must be measured rather than estimated, as the error contained in the estimation is significant enough to bring the whole thing down to guessing.

Just like with the power supply itself, if you put a oscilloscope on the board and measure the consumption of the cpus you will find it likes at a much lower level than common wives tales, estimates and ratings dictate.

So, although their max estimated power consumption is 91.6 watts, they are probably using somewhat less than that, even at full load?

larva

10-09-03, 01:53 AM

Doesn't that make sense considering the measured total consumption level? You know the total power is only 233W, so what is the chance the cpus really draw 91.6W apiece?

To hopefully make clear what I hope is obvious, consider this: that 91.6W would be the output of another switching power supply on the motherboard, with its own attendant power loss. While higher in efficiency than the main supply there is no perfect machine, and efficiency is always less than 100.

cmcquistion

10-09-03, 09:28 AM

So, you're saying that there is power loss, due to efficiency, at the motherboard level, also?

Are you also saying that the motherboard may be drawing that 91.6 watts per CPU, but some of that wattage is lost, due to the efficiency of the conversion? Therefore, the CPU's are using some amount less than that 91.6 watts? (maybe 70% or something)

I just want to make sure I understand, fully.

larva

10-09-03, 12:32 PM

No, what I am saying is the exact same factors apply at the motherboard level, as modern motherboards employ a switching power supply to derive regulated Vcores less than 3.3V. The quality of parts is higher here due to the economy that the smaller scale of the supply affords, but it's still a switching power supply with it's own efficiency value less than 100%.

Lets assume (making an ass of u and me) that the cpu's do draw 91.6W each. Lets further assume that the motherboard's regulator achieves a 90% efficiency. (I'm going to guess that motherboard regulators come in at 85-95% efficiency, but understand I am just guessing here). That would imply that the motherboard has to draw (91.6/.9)=101.8W. Times two. 203.6W. Leaving 30W for the entire rest of the the machine.

Do you really think the cpu's draw that much? ;)

cmcquistion

10-09-03, 12:34 PM

Ok, so you're saying that the motherboard probably draws 91.6 watts per CPU, but the CPU's, themselves, aren't really using that much.

dustybyrd

10-09-03, 01:20 PM

Ok, so you're saying that the motherboard probably draws 91.6 watts per CPU, but the CPU's, themselves, aren't really using that much.

if that were true, that would still only leave 20 more watts for the entire rest of the system...or a total of 50 watts for the rest of the system...

and that's still not likely...the rest of the system probably would need more than 50 watts...2 HD's, 2 optical drives, video card (which we already know draws 30 more watts when 3dmark is running), RAM, mb....

therefore, the cpu's aren't drawing that much...probably 70 watts each...even under load...or it just wouldn't add up...

the 91 watts must be based on the max possible albeit for a brief, unlikely moment

cmcquistion

10-09-03, 01:23 PM

Originally posted by dustybyrd
if that were true, that would still only leave 20 more watts for the entire rest of the system...or a total of 50 watts for the rest of the system...

and that's still not likely...the rest of the system probably would need more than 50 watts...2 HD's, 2 optical drives, video card (which we already know draws 30 more watts when 3dmark is running), RAM, mb....

therefore, the cpu's aren't drawing that much...probably 70 watts each...even under load...or it just wouldn't add up...

the 91 watts must be based on the max possible albeit for a brief, unlikely moment

Good point. That is probably the case.

It isn't really all that important for me to know concrete numbers, it was just something that seemed fishy to me, when I used the calculated numbers.

You and larva have helped me see the light, though...

dustybyrd

10-09-03, 01:58 PM

check this link out:

http://forums.silentpcreview.com//viewtopic.php?t=3321

using an ammeter to measure the current requirements on each line of a dual AMD...

note the chips are 1400's...and that power calculator suggests ~50-60 watts each on them...

cmcquistion

10-09-03, 02:08 PM

So my figures pretty much jive with those figures (except my system is using more power due to faster CPU's and a GF4.)

Notice his system, which is not overclocked and doesn't include a high power video card, is still drawing 27 amps on the 5V line. Add faster CPU's and/or a power-hungry video card, and you quickly add up to more amps off that 5V, then any 300 watt PSU offers.

My Fortron FSP300, which is the best 300 watt PSU I have EVER used (some testing has revealed that this PSU supplies as much power as many 400-450 watt PSU's), only offers 30 amps on the 5V line. This is better than most. Of course, larva's earlier point still stands, however. Most PSU manufacturer's supplies ratings are unusable, really, so even if a PSU claims to deliver X amps on the 5V line, that doesn't really mean it can deliver.

Also, from his article:

The total AC power draw of the system was measured with a Kill-a-Watt power meter.

At idle, it measured 210W.
At maximum, it was 250W
If one assumes 70% PSU AC/DC efficiency, this translates to

At idle, 147W DC power delivered to components.
At maximum, 175W DC power delivered to components.

Those figures don't totally jive with his later measurements with the amp-meter.

Results: (in order of 1] CD/DVD, 2] hdd's and 3] motherboard)

3.3V: 0A / 0A / 2.5A: total = 2.5 A / 8.25 watts
5V: 1A / 1A / 27A: total = 29 A / 145 watts
12V: 1.7A / 1A / 1.5A: total = 4.2 A / 50.4 watts

Total power consumption: 203.65 watts

His measurements with the amp-meter show a usage of about 203 watts and this is AFTER the loss due to efficiency of the PSU. This would seem to indicate his PSU efficiency may be about 81%. Of course, this could just be normal deviation, because of the different measurement devices used.

dustybyrd

10-09-03, 03:29 PM

Those figures don't totally jive with his later measurements with the amp-meter.

His measurements with the amp-meter show a usage of about 203 watts and this is AFTER the loss due to efficiency of the PSU. This would seem to indicate his PSU efficiency may be about 81%. Of course, this could just be normal deviation, because of the different measurement devices used. [/B]

they might still jibe because...according to that silentpcreview web site...when they did psu testing (pretty thorough i might add)...they found that for the antec tp 550 that efficiency was highest in the middle power ranges...much lower at low power and lower at high draw....

and they found that for multiple psu's

cmcquistion

10-09-03, 03:30 PM

Originally posted by dustybyrd
they might still jibe because...according to that silentpcreview web site...when they did psu testing (pretty thorough i might add)...they found that for the antec tp 550 that efficiency was highest in the middle power ranges...much lower at low power and lower at high draw....

and they found that for multiple psu's

Yeah, I thought that the PSU's efficiency might work that way. Thanks for digging that up!

larva

10-09-03, 04:12 PM

Originally posted by cmcquistion
Ok, so you're saying that the motherboard probably draws 91.6 watts per CPU, but the CPU's, themselves, aren't really using that much.

Sorry, got a new hard drive and I was intalling/testing it. I'll post a mini-review of the Hitachi/IBM Deskstar 7k250 SATA drive in the storage section.

I'm saying I don't see it possible for the cpus to be drawing that much. Power supply efficiency does vary and lots of other factors conspire to complete the total picture, but measured accurately under load I don't feel the cpu's will draw 91.6W each in practice.

cmcquistion

10-09-03, 04:18 PM

Originally posted by larva
I'm saying I don't see it possible for the cpus to be drawing that much. Power supply efficiency does vary and lots of other factors conspire to complete the total picture, but measured accurately under load I don't feel the cpu's will draw 91.6W each in practice.

ok.

L337 M33P

10-09-03, 04:33 PM

Here's something you might find useful: It is a graph of a switch mode PSU's efficiency with load.

http://www.elecdesign.com/Files/29/2638/Figure_01.gif

It is a high-efficiency DC-DC converter though.

silentmc

10-11-03, 12:11 PM

larva obviously has a great deal of knowledge & understanding about PSUs and power ratings. My kudos. Your participation at http://forums.silentpcreview.com/ would be much welcome by the members there!

cmcquistion, on the other hand, to brand us at http://www.silentpcreview.com as "the evil forum" is plain silly. We're the ones who showed you the Kill-a-Watt, to educate you enough to even begin asking these questions about PSUs in the first place. The "CPU Power calculator" you refer to as "my" is in fact a utility created by SPCR member Kostik; such lack of common decency & respect in not acknowledging this fact!

For the enlightment of oc members, please see:
http://forums.silentpcreview.com/viewtopic.php?t=7593

EDIT: Post edited for brevity.

cmcquistion

10-11-03, 12:29 PM

Hello Mike.

For an exhibition of your "good manners" at the SilentPCReview forums, please read this thread:

http://forums.silentpcreview.com/viewtopic.php?t=6731

There are countless other examples of the extreme hostility and rudeness of the members, there. Your forum is ripe with flaming and posturing.

This kind of hostility, rudeness, name-calling, and flat out lying is why I lightly refer to your forum as "the evil forum".

Is there good information there? Absolutely. It was the only place that I could find actual user information of the Kill-A-Watt device, when used with computers. Unfortunately, the conduct of the members in that forum is horrible to put it mildly.

*EDIT*

EDIT: Post edited for brevity.
I wish I had quoted your original post, before you edited it, which included some obvious flaming. Flames (personal attacks) are never allowed in this forum, unlike some other forums.