Rationalize the 500w vs 600w PSU

mikegirvan

n00b
Joined
Nov 15, 2005
Messages
1
Someone explain this for me.

From what i've read most new computers with a reputable powersupply will be ok with a 450w PSU. Some say get a 500w PSU for comfort, this way you'll be ok with a dual core + SLI machine.....

With a avg 3-5 year warranty on PSU's why not just spend the extra $20 dollars(on avg) and get the reputable 600w PSU, this leaves you plenty of breathing room should technological advances require more power down the road.....

This all stems from me building my new system now- Its going to be dual core with a GTX 7800 card but may potentially turn into a SLI setup in the future.....it just seems to make sense to get the extra 100watts of power for an additional 20 bucks then find out in a year or two an upgrade will be needed due to increased power demands....

Thoughts?
 
Depends on what rails the added capacity is on....but very few people actually use 500w's total out of the PSU. Most of the time your draw is going to be heavy on one rail.....modern systems it is the 12v...so 12v amps is far more important than a useless wattage rating.
 
Three things here:

1) Watts are not all that meaningfull to measure a power supplies actual output, amps are. You can buy a "600" watt el-cheapo PSU and it'd be lucky to muster 300 watts actual output on a good day.

2) Overkilling on the PSU has a drawback in that the lower load you put on the PSU the efficiency will go down, which creates more heat in the PSU and a bigger electric bill unnecessarily. e.g. a 400 watt PSU with a 350 watt load will work more efficiently than putting a 200 watt load on the same PSU.

3) PC's do not use as much power as many people think, money invested in a higher quality PSU goes a lot further than just getting the highest wattage you can. I have a lot of machines at home calculated how much actual power they use (with a watt meter). My workstation (XP 3200+, 9800 PRO, 2x 512 MB, 2x 7200 RPM, 2x DVD burners) uses less than 120 watts at idle, and under full load gaming, it draws less than 275. SLI with some heavy duty video cards will need a little more, but not as much as you'd think.
 
well my question stems from my inability to choose the Seasonic S12 500w or 600w PSU as my next purchase.

I plan to have a quality gaming machine but im still unsure of the return of investment if i were to go with a 500w or 600w PSU from this manufacturer.....


600 just sounds logical to me in the longterm - upgradibility, etc..


/me shrugs
 
mikegirvan said:
Someone explain this for me.


With a avg 3-5 year warranty on PSU's why not just spend the extra $20 dollars(on avg) and get the reputable 600w PSU, this leaves you plenty of breathing room should technological advances require more power down the road.....

Thoughts?

When I went psu shopping the jump from 500 to 600 watts was quite a bit more than $20.
 
I agree with you.

My current file server is hooked up with a cheapie Allied 500w psu...its powering:
P4 3.06northwoood w/Intel 865 on board vga
2 raid cards
1 gig nic
1 dvd burner
11 hdd

According to a psu calculator I'de need another 50 watts for another raid card with 2 more hdds...
Since its using the onboard video do have plans to make it my 2nd gaming machine (most likely with my AIW 9800pro in the TV computer) so I know I'de need more watts for that...so I got myself a Silverstone 650W Zeus psu for $123 at Fry's.
 
Not all PSUs are created equal - even within the same wattage class. Modern systems draw most of their power from the +12V rail - but there still are a few 600W power supply units which have weaker +12V rail(s) than some 500W or even 400W units.
 
So as the end all, would this mean that a Seasonic 600W is actually not a good buy if you're not running 10 HDD, 2 physical CPUs, and SLi?

I'm in the market for a new PSU, and was going to go with that one, however, would it be more "efficient" if I were to go with the 500w version, or possibly the 520W OCZ modstream? You can see my Sig to get an idea what I'm runing. The only difference is I'll be geting a dual core AMD and 2 Sata drives (1 raptor), and DO plan to OC. I don't want to fall in the same boat I'm in now. I want a nice quality PSU. I'd rather modular, but hey, "Can't have your cake and eat it too . . ."

Any suggestions?
 
so we can close this thread by saying...I dont need the 600w, it wont be as efficient- just get the 500w

?
 
Lots of misinformation in this thread. What's most depressing is that in the case of the Seasonic S12, there's good information out there. So rather than using sweeping generalizations about how power supplies respond to less than maximum load, one can actually see real results.
http://www.silentpcreview.com/article247-page1.html

Turns out the efficiency numbers from that review were bogus, however. Thankfully, they've been recalculated :cool:..
http://www.silentpcreview.com/article276-page1.html

And for the record, I bought the 600W version simply because I got a good price on it. With normal pricing, I'd say 500W is plenty sufficient. The system below draws a bit under 300W from the outlet under maximum load. About 100W of that is just the monitor and speakers, though. So as you can see, there's a ridiculous amount of headroom in this power supply, and likewise for the 500W.
 
picasso1850 said:
so we can close this thread by saying...I dont need the 600w, it wont be as efficient- just get the 500w

?

No.. You have that b-ass-ackwards. The 600W will be more efficient.

Power supplies are more efficient at lower loads. For example: A 600W with a 300W load may be 80% efficienct. A 500W with a 300W load may be 75% efficient.

Better example: How did PCP&C make a 1kW PSU? They took the 850W and made it more efficient. So now, instead of having > 70% efficiency up to 850W, you now have > 70% efficiency up to 1000W.

Get it?
 
one would be hard pressed to spot any operational difference in the efficiency values as well (either noise or heat) most of the time, and certainly not in the electric bill :p

a supply that has "over capacity" is in effect one that will be able to run cooler for the load on it, with possibly a slightly lower efficiency, but the real advantage is in the ability to deal with adverse conditions since its not operating near its rated capacity (exceptional raised ambient temperature, brownouts and surges, very heavy transient loads and peak draws) and of course the upgrade path

it used to be argued that such behmoths where too loud, but a good one will have a thermostatically controlled fan, and more often than not a larger heatsink (interface area) than its smaller sibling allowing it to run cooler for a given load at a given ambient with the same volume of air per second

that generally overcomes any minor effciency considerations when it comes to the actual ability to dump the heat
slightly less effcient AC to DC conversion with a more effcient means to dump the heat being largely a dead heat on the merry-go-round when it comes to selecting a supply at least for the individual, for an enterprise buying a few hundered then the electric bill weights in
(infrastructure for thermal density - AC and industrial chillers, KVA/UPS support, actual consumption)
what it really means is that a PSU manufacturer can reduce the size of the heatsink, lowering the cost and shipping and look good to the giovernment
for the indivudual end user there is no great impact everything else being equal

it might also have a louder fan than it siblings which is why we have (very few) competent reviewers
like Mike @ SPCR, jonnyGURU @ SLCentral, Lee Garbut @ Various and Oleg Artamonov @ Xbit
who provide overall context, its far too easy to blow one spec far out of proportion to its real world impact
<cough > THG </cough>

of course thats comparing apples to apples in a given wattage class which of course isnt the real case, many 500watts not being equal to 350watts of another make, again why we need good reviewers, unfortunately we are over run with a horde of cheerleader dupes posing as reviewers
they are fine for esthetic reviews but lack an understanding of what the specs are and mean ,
how they are interrelated to real world performance and how to actually test them
worse they spread disinformation based on thier limited understanding and flawed proceedures
 
Ice Czar said:
one would be hard pressed to spot any operational difference in the efficiency values as well (either noise or heat) most of the time, and certainly not in the electric bill :p

a supply that has "over capacity" is in effect one that will be able to run cooler for the load on it, with possibly a slightly lower efficiency, but the real advantage is in the ability to deal with adverse conditions since its not operating near its rated capacity (exceptional raised ambient temperature, brownouts and surges, very heavy transient loads and peak draws) and of course the upgrade path

it used to be argued that such behmoths where too loud, but a good one will have a thermostatically controlled fan, and more often than not a larger heatsink (interface area) than its smaller sibling allowing it to run cooler for a given load at a given ambient with the same volume of air per second

that generally overcomes any minor effciency considerations when it comes to the actual ability to dump the heat
slightly less effcient AC to DC conversion with a more effcient means to dump the heat being largely a dead heat on the merry-go-round when it comes to selecting a supply at least for the individual, for an enterprise buying a few hundered then the electric bill weights in
(infrastructure for thermal density - AC and industrial chillers, KVA/UPS support, actual consumption)
what it really means is that a PSU manufacturer can reduce the size of the heatsink, lowering the cost and shipping and look good to the giovernment
for the indivudual end user there is no great impact everything else being equal

it might also have a louder fan than it siblings which is why we have (very few) competent reviewers
like Mike @ SPCR, jonnyGURU @ SLCentral, Lee Garbut @ Various and Oleg Artamonov @ Xbit
who provide overall context, its far too easy to blow one spec far out of proportion to its real world impact
<cough > THG </cough>

of course thats comparing apples to apples in a given wattage class which of course isnt the real case, many 500watts not being equal to 350watts of another make, again why we need good reviewers, unfortunately we are over run with a horde of cheerleader dupes posing as reviewers
they are fine for esthetic reviews but lack an understanding of what the specs are and mean ,
how they are interrelated to real world performance and how to actually test them
worse they spread disinformation based on thier limited understanding and flawed proceedures


All that knowledge you just dropped is giving me a headache. :D Also quoted for truth.
 
Back
Top