Seasonic 500W

[Matthias99]

Originally posted by: Luckyboy1
Intel EE rep said first that the Prescott. 3.2 GHz CPU used 120 amps. Then later, he admitted to 130 amps. Then add to that people sick enough to Peltier the thing and run higher voltages and 140 amps is not unheard of in that situation.

I'm gonna assume you mean "watts" there and not "amps" (unless maybe they meant amps at 1.6V or whatever CPU supply voltage is on the Prescott, but even that sounds high). TDP for the Prescotts is around 125W, but Intel's TDP numbers are not the 100% theoretical worst-case. Overclocking will also significantly up the power draw, especially if supply voltage is increased.

The vast majority of the CPU's power draw is from the +12V rail.

ABIT confirmed the later numbers and said get something that can put out 300 watts total on those combined rails to run an ABIT IC-7-MAX3 mommaboard with that CPU and an 800XTPE along with all the hard drives and Audigy 2 ZS card.

English, please.

You'd have to try REAL HARD to get anywhere near 180W of load on the +3.3V and +5V rails, let alone 300W. The kind of system you're describing might hit about 300W total, not 300W on the +3.3/+5V rails.

 

[Operandi]

Originally posted by: Luckyboy1
I don't have an idustrail strength amp meter. I simply don't need to read all that dirarrhea! I will, but wow! what a waste of space!

The 478 pin board you show is a Northwood, not a Prescoot setup! Ding!

So what? There is a 3.8GHZ Pentium D system that draws 183 watts total, not just the CPU the whole system!

Originally posted by: Luckyboy1
It's got a lousy 6 pixel pipeline video card which draws much less, not just in +12 volt power than mine, but other values as well! Ding!

The most power hungry card they test is the 6800GT which is only a bit short of what is a available today in terms of power draw. If bothered to look you can see what it draws in relation to the other cards.

Originally posted by: Luckyboy1
At this point, I'll stop reading your useless link and defer back to the Electrical Engineers who made my gear in the first place!

Ok, so your content with having absoultly no understanding of power supplies and power draw in general; fine. But for the sake of everyone else on this forum please spare us from your posts.

 

[Fullmetal Chocobo]

Okay. Using a Seasonic Power Angel to measure, my system pulls is at 445 watts. And my system is about as power hungry as a machine can get. That measurement was taking while running DPAD (100% processor usage on all four "cores"), defragging all 8 hard drives, and running 3DMark03 for the 6800 GT. So the chances of just a processor pulling 300 watts is pretty impossible I'd say.

 

[JEDIYoda]

well ya know we those of us who are forum members have seen Luckyboy1 owned more than once...
Also note we all have various brands that we like but all Luckyboy1 pushes are OCZ PSU`s...
At least some of us also push...Fortron,,,Seasonic...Zippy...Enhance...Enermax...


take this thread...notice how he completely avioded Zepper...hmmm

http://forums.anandtech.com/messageview.aspx?catid=37&threadid=1824711

re. original post,

Wrong on point 1! If you isolate the intake and exhaust air for the PSU, it doesn't matter where they are located: top, bottom front, back or suspended on springs in the physiical center of the case (assuming empty space exists there. Just because most cases that do a bottom mount for the PSU also happen to isolate its airflow at least to some degree, is not proof positive that the bottom-mount PSU is the way to go. Take a look at the Arctic Cooling Silentium cases before writing the mass off.

Wrong on point 2! There are many useful items besides optical drives that reside in 5.25" bays. I have a mobile drive rack in one and fan controller (with a HD tucked in behind it) in another.

Iffy on point 3. I usually see no point at all in side mounted fans.

Iffy on point 4. Case designs with motherboard drawers would have to be made substantially wider to allow for 120mm fans in the rear thus incurring extra cost for the case material (and wasted but for the 120mm fans) and extra shipping as the basic carton would exceed the 84" limit for many standard shipping rates. But a 92mm mounts would fit most drawered cases - that definitely should be done. From that point improvement could be made easily if needed.

So lets tally the score: Point 1 - 0, point 2 - 0, point 3 - 0.5, point 4 - 0.5. Total = 1 out of a possible 4 = 25%... !AHHNNNH! - You flunked!

.bh.

No response interesting..lolol

 

[Matthias99]

Originally posted by: Fullmetal Chocobo
Okay. Using a Seasonic Power Angel to measure, my system pulls is at 445 watts. And my system is about as power hungry as a machine can get. That measurement was taking while running DPAD (100% processor usage on all four "cores"), defragging all 8 hard drives, and running 3DMark03 for the 6800 GT. So the chances of just a processor pulling 300 watts is pretty impossible I'd say.

Also note that you're measuring power usage at the wall, so that's not taking the PSU's efficiency into account. With an 80% efficient PSU, your actual system draw would be ~350W. With a 70% efficient PSU, your system would only really be using ~310W.

 

[Fullmetal Chocobo]

Originally posted by: Matthias99

Originally posted by: Fullmetal Chocobo
Okay. Using a Seasonic Power Angel to measure, my system pulls is at 445 watts. And my system is about as power hungry as a machine can get. That measurement was taking while running DPAD (100% processor usage on all four "cores"), defragging all 8 hard drives, and running 3DMark03 for the 6800 GT. So the chances of just a processor pulling 300 watts is pretty impossible I'd say.

Also note that you're measuring power usage at the wall, so that's not taking the PSU's efficiency into account. With an 80% efficient PSU, your actual system draw would be ~350W. With a 70% efficient PSU, your system would only really be using ~310W.

I know, and it only goes to prove my point more. I'm running two Seasonic PSUs, so they are fairly efficient, which is why I didn't bother to make that point as well. Also, when I do run on just the 600w Seasonic, it isn't the 3.3v or 5v rails that suffer, it's the +12V.

 

[furballi]

Correct. The most accurate power measurement is with an AC ammeter connected in series to the PC. The accuracy is within +/-3%. Use an efficiency factor of 80% to derive the actual power used by the PC. The average AMD rig with a high-end video card will pull about 250 watts. Add another 75 watts if you're running an Intel platform.

 

[biostud]

Originally posted by: jack bauer
Will a seasonic s12 500W be enough for:

A8N32-SLI, 2x 7900GT, AMD opteron 170, 2 fans, one HD, no soundcard, 2 DVDs

Yes

And your PSU will not be the most noisy in your computer.

 

[ribbon13]

Originally posted by: Matthias99
I'm gonna assume you mean "watts" there and not "amps" (unless maybe they meant amps at 1.6V or whatever CPU supply voltage is on the Prescott, but even that sounds high). TDP for the Prescotts is around 125W, but Intel's TDP numbers are not the 100% theoretical worst-case. Overclocking will also significantly up the power draw, especially if supply voltage is increased.

Check bottom spec

Anyway, S12-600 is very nice, but if you truly require massive amounts of DC, you'll pay for it in your power bill but that will definately be sufficient amounts of power. I'd stick with the S12-600 given the type of power draw you'll be running.

 

[bigpow]

Originally posted by: fire400
unless you want a silent PC, go with the seasonic

I don't know how silent this setup would be in accord with how effective you would cancle out noise with the seasonic:

A8N32-SLI, 2x 7900GT, AMD opteron 170, 2 fans, one HD, no soundcard, 2 DVDs

ROFL

seasonic with this rig: RIP OFF

HUH?!
I think my S12-600 is quiet enough, even quieter than the Scythe 1200RPM fan that came with the Ninja Plus.

I wouldn't worry about Seasonic PSU noise in that setup - most likely the pair of 7900GT will be the loudest

What other cheaper PSU you would recommend, that's: 1) not a rip off 2) quieter than Seasonic?

 

[Capt Caveman]

The fan on my Seasonic S12 500 never goes above 900rpm and I can only hear it if I put my head against the back of my case. OP - I was able to run the system in my sig w the S12 500 w/out any issues. And I know several folks in the Video forum that run 7800GT SLI(soon to be 7900GT SLI) w/out any issues.

But if you're concerned, spending an additional $30 or so for the 600 wouldn't be a bad thing.

 

[Luckyboy1]

Whoops! It is watts of course and not amps!

Look, guys, remember just a couple of years ago when the ABIT IC-7-MAX3 came out? Before that, few mommaboards if any had any kind of cooling for their power section. Then you started seeing at least passive cooling in the form of heatsinks. Then they whipped fans on them. Was this a sudden understanding that they were wrong about cooling that section of the mommaboard all those years? Of course not! Today, active cooling of the power section of mommaboards is while not universal, alot more common. Why? Higher power needs of the board itself. Combine that with higher power demands of the CPU and video cards, much less all the other stuff that goes along with it and nobody should be surprised that in most gaming PC's a $50.00 power supply just won't make it. Even if the PC "runs" on that power, how is it hobbling performance?

All I can do is pass along my real world results and when I finally went to a stout power supply, the chips ran cooler and I could clock further and all kinds of transient funny things stopped happening. Beating up on me won't change that regardless of typos!

 

[Matthias99]

Originally posted by: ribbon13

Originally posted by: Matthias99
I'm gonna assume you mean "watts" there and not "amps" (unless maybe they meant amps at 1.6V or whatever CPU supply voltage is on the Prescott, but even that sounds high). TDP for the Prescotts is around 125W, but Intel's TDP numbers are not the 100% theoretical worst-case. Overclocking will also significantly up the power draw, especially if supply voltage is increased.

Check bottom spec

70A@1.3V = ~90W. 😛 I'm not exactly sure what voltage Prescotts use, but unless it's sub-1V, 140A is just not happening.

 

[Luckyboy1]

You are quoting original release specs that were well... marketing over reality! Over time, the truth came out and yes, you bump the voltages enough and it is possible to reach those values. It was known that a 3.2, 478 pin Prescott was hobbled by bandwidth problems until you reached about 280 on the multiplier in 1:1 mode, where it counts! I'm far beyond that now and yes, the results are nice, but they were far from "worth it".

Of course to reach 4.4 GHz rock stable in an 8 hour torture tests... and I run mine for 72 hours in case you wondered... you'll have to get a 24 amp Peltier and the power supply to run it... a dedicated, sperate power supply because at the time, there was no single power supply for PC's sold that met the need.... you'll need to water cool the Northbridge, actively cool with air the Southbridge and at least put a heatsink on just about everything else on the motherboard. All this creats tremendous heat and in the end, was simply a great way to prove cooling and power principles and what is or isn't needed and when and when something is not of benefit, but as far as a practical matter, it was FUBAR!

Still, even if you are not climbing way up on the list in the bios to find the insaine voltages you are going to run at, with today's technology, we are more power hungry than ever. The mommaboards ARE designed with cooling for the power circuits and ignoring this design consideration and the power needs that comes with it is hiding your collective heads in the sand.

 

[t3h l337 n3wb]

What's up with the term "mommaboard"? Motherboard is much better sounding 😛

Back on topic, your original argument was just that the +3.3V and +5V loads on the S12-500 weren't high enough. It doesn't matter that components keep needing more power, because they draw most of it off the +12V rail now. Older parts used the +3.3V and +5V rails more, but the ratings on them really aren't very important now.

 

[Howard]

Originally posted by: Luckyboy1
You are quoting original release specs that were well... marketing over reality! Over time, the truth came out and yes, you bump the voltages enough and it is possible to reach those values. It was known that a 3.2, 478 pin Prescott was hobbled by bandwidth problems until you reached about 280 on the multiplier in 1:1 mode, where it counts! I'm far beyond that now and yes, the results are nice, but they were far from "worth it".

Of course to reach 4.4 GHz rock stable in an 8 hour torture tests... and I run mine for 72 hours in case you wondered... you'll have to get a 24 amp Peltier and the power supply to run it... a dedicated, sperate power supply because at the time, there was no single power supply for PC's sold that met the need.... you'll need to water cool the Northbridge, actively cool with air the Southbridge and at least put a heatsink on just about everything else on the motherboard. All this creats tremendous heat and in the end, was simply a great way to prove cooling and power principles and what is or isn't needed and when and when something is not of benefit, but as far as a practical matter, it was FUBAR!

Still, even if you are not climbing way up on the list in the bios to find the insaine voltages you are going to run at, with today's technology, we are more power hungry than ever. The mommaboards ARE designed with cooling for the power circuits and ignoring this design consideration and the power needs that comes with it is hiding your collective heads in the sand.

You're good at diversion.

 

[JEDIYoda]

actually I have had my P4 3.2 EE rock solid stable with just air cooling at 4.2 for months at a time.
I have it all upto 4.4.......stable but things like that have always made me a bit nervous plus you tend to get dizzy when your up there in that rarified air.....lol

 

[Luckyboy1]

Originally posted by: JEDIYoda
actually I have had my P4 3.2 EE rock solid stable with just air cooling at 4.2 for months at a time.
I have it all upto 4.4.......stable but things like that have always made me a bit nervous plus you tend to get dizzy when your up there in that rarified air.....lol

A man after my heat and soul! 🙂

Not for getting that far overclocking, but for being smart enough to know that if he backed off a bit now, he could keep that clock just about forever! :beer: