Why 80 PLUS® is Irrelevant to You When Buying a PSU @ [H]

FrgMstr

Just Plain Mean
Staff member
Joined
May 18, 1997
Messages
55,534
Why 80 PLUS® is Irrelevant to You When Buying a PSU - A HardOCP editorial about the 80 Plus® program and how it is changing the computer power supplies you are buying. Is this good, bad, or ugly, and should you care? Did you know that you paid for that 80Plus rating on your shiny new PSU? Certainly PSU efficiency is a great thing, but what about the rating system?
 
nice. very much needed to be said. Hopefully kyle emailed copies of this to certain people.

and high kudos to H for failing the cheaters. needed to be done.

to add to the cost saving consideration mentioned one more may be need to address depending on the user. if you have to cool your PC room (on any scale) this does make a far greater difference. simply put you have not only to account for the wasted energy but the cost in energy of removing the waste heat.
 
Wow, great job Paul/Kyle, you opened a nice 55 gallon drum of whoop ass on this one. Need more guys with [H]ardbawls like you here to kick the industry straight.
 
Agreed, good job putting this together and getting it out there. Hopefully it (the article) gets picked up and spread around the 'net to inform more people.
 
I'm glad I don't blindly follow these 80 plus certifications. Luckily I normally stick with Corsair or Seasonic and sometimes Antec, if the price is right.
 
Power here is closer to 14c kw/h so the savings is much higher so if I was running a high power draw computer it would be worth it to jump up to a correctly graded higher efficiency supply.
Most of the time my computer is drawing around 65W so not worth upgrading from the current 80plus.
 
I only buy power supplies that have passed rigorous testing. I plan on sticking to that rule and glad I can count on [H] to provide those results. Thanks! Great article and very well thought out.
 
I deleted my first post, realizing it was irrelevant.

Paul, Kyle, one other thing I'd love to see pointed out in an article. It can be far more efficient to run a 550w power supply that is 80Plus Bronze than, say a 1000w unit that is 80Plus Gold, just because someone may not be loading the unit up to the wattage level that is measured for the 80Plus certifications.

I think a lot of people make the mistake to buy way more headroom than they need on a power supply, and then it costs them efficiency because they're using half, or a third of what it is capable of. It would be great to make some comparisons about what power draw at the wall is for entry, mid, high, and uber-level systems really are, because a lot of people are shooting in the dark.

I'm sure I could easily use a power supply rated 100w less than what I have, provided it's a quality unit (I would have, but this one was on sale for less :)). Possibly even less than that, but it would be great for all of us to know just how far overboard we might be going in our quest for the best.
 
Last edited:
Curious side question, Obviously power supplies aren't linearly efficient correct?

So where are most of them most efficient? And shouldn't that matter more on what size PSU you buy? Trying to hit the best efficiency average load?

Still a great effort on HardOCPs part, and really sticking up for the consumer and not letting advertising change that.
 
Comparisons at 8 cents a kWh? Really?

8 cents is well under the national average for power pricing. Even the average is under what anyone who is subject to tiered billing will be paying, since there is a significant amount of 'baseline power' that uses power from the lower tiers no matter what (lighting, cooking, refrigerator / freezer, etc...) which will shoot you into a higher tier. My personal power cost at the tier where a high efficiency PSU would save me money from is between 30 and 35 cents, about FOUR times the value you used. My parent's, who live 500 miles away and have a different power company are also in the 30-35 cents range.

There are also several power companies that charge separate fees for power generation costs and power transmission costs. Mine does not, it rolls everything into one number, but my parents thought they were paying like 15-16 cents a kWh at their highest tier, but I pointed out that their bill shows generation / transmission separately and both are per kWh and tiered. They are really paying over 30 cents a kWh. So how many other people are reading their bill wrong?

If you are going to show graphs from the standpoint of a ludicrous underestimate of the average power cost, you should also take a data point from someone at the higher end of the average.

To put this into some perspective let’s briefly look at an example of pricing with the closest thing we have to a 500 watt 80 PLUS® vs. 80 PLUS® Platinum scenario. Currently, the 80 PLUS® Platinum Kingwin LZP-550 (an excellent power supply by the way) is available for $159.99 while the 80 PLUS® Silverstone ST-50F 500W is available for $49.99 and Corsair CX500 is available for $39.99. Under the scenarios we constructed above, where your electricity cost savings would be the greatest among 500 watt units, it will take a very long time to make back the price difference when buying the LZP-550.

Well, dur. Of course it is when you use deflated power costs.

Let's do the same math with the Seasonic 560W 80+ gold on newegg right now for <$120. At my actual power costs (33cents) the difference between 80+ and 80+ gold is a little under $40 a year at the 20% usage scenario... the lowest usage scenario presented in the article. I've mapped my own usage via a kill-a-watt and it's actually not far off from the 20% usage scenario.

Yes, the article says this will vary for users, but it presents a pretty rose colored glasses set of scenarios. My last PSU was in use for 5 or 6 years, and when you multiply your scenarios by 4 to get a more realistic power cost, then by 3 - 5 years, you can easily justify higher efficiency on cost alone.

Platinum is a silly tier, due to how PSU manufacturers are meeting the spec (mostly just slight tweaks and de-rating a gold design). But 80+ gold level PSUs are often the cheapest in the long term, even compared with Bronze.

----------

In the article, you also state that 80+ cert is largely a marketing tool, and this is why MFRs cheat the specs.

But, if review sites like HardOCP continue giving awards and recommends to power supply MFRs who cheat the specs, of course it will largely be a marketing tool. It will encourge more MFRs to start cheating the specs, lying to customers, and getting accolades as well.

On the other hand, if sites were to slam the MFRs and call them on the table for lying to consumers, it will highlight the issue, and reward those MFRs who do not cheat. Seems like what a review site should be doing. And the irony of giving recommends / awards to supplies that do not meet advertised specifications sounds like the kind of humorous news article I'd see Steve posting on the front page.

You are pretty brutal to supplies that do not meet other advertised recommendations, it seems unjust to be lax about the 80+ cert just because power in your area is apparently cheaper than pretty much anywhere else in the country.
 
Comparisons at 8 cents a kWh? Really?

I understand that you live in California and have to pay for your states issues which means you pay significantly more than the rest of us, and you do not feel that the rest of the country paying less than you is fair. Now if it helps, the national retail average cost of electricity in 2010 according to the DOE was 9.88c kw/h. You can re-figure the numbers based on the national average if you would like, but the 1.88c difference is not going to change the point that was being illustrated.

Now if you had read the article without being so emotionally invested in the cost of your electricity, you perhaps would have understood that this was being used as a general frame work to illustrate a point as was stated as the end of page 2:

We are not going to assign solid numbers to the following explanations because so many variables can be introduced. While PSU pricing, usage scenarios, and the cost of electricity are not static, a trend we clearly see is that the higher the level of 80 PLUS® certification on a PSU the more expensive the unit is at its output level. Additionally, while the time to recoup the cost is going to vary, the general trend is that due to the higher pricing of the higher ranked 80 PLUS® units your time to recoup that premium is going to be significantly longer unless you are in a rather narrow niche of people who use their system 24/7 at significant load levels, which the majority of consumers and even enthusiasts are likely not.

Take it as what it is, a general framework to illustrate a point using numbers closer to the national average than what parts of California pay.

In the article, you also state that 80+ cert is largely a marketing tool, and this is why MFRs cheat the specs.

But, if review sites like HardOCP continue giving awards and recommends to power supply MFRs who cheat the specs, of course it will largely be a marketing tool. It will encourge more MFRs to start cheating the specs, lying to customers, and getting accolades as well.

On the other hand, if sites were to slam the MFRs and call them on the table for lying to consumers, it will highlight the issue, and reward those MFRs who do not cheat. Seems like what a review site should be doing. And the irony of giving recommends / awards to supplies that do not meet advertised specifications sounds like the kind of humorous news article I'd see Steve posting on the front page.

You are pretty brutal to supplies that do not meet other advertised recommendations, it seems unjust to be lax about the 80+ cert just because power in your area is apparently cheaper than pretty much anywhere else in the country.

Perhaps in the future you should read all of the article ;) From page three:

We recently added our own version of 80 PLUS® certification testing to our PSU reviews and will continue to run ECOS' 80 PLUS® certification loads in our reviews. We started doing this because of the intense corporate focus on 80 PLUS® certifications, not because we find these all that useful. If a power supply comes in that misses 80 PLUS® certification numbers by a wide margin we will fail that unit.
 
Last edited:
Good stuff as always with PSU reviewing at [H], wish I knew about this site before I bought the PSU I did.
 
If you care about energy efficiency, you'll buy a system that uses less energy first, THEN worry about the power supply efficiency.

Yeah, I thought about replacing the PSU in my media center PC with an ultra-high-efficiency one. It would have cost $250. To save, what, 5 Watts? (Admittedly, it also would have been quieter.) But I will get a lot more bang for the buck by upgrading the video card. Same GPU horsepower, newer process, lower power, more features. So I'll save 20+ Watts for $150.

Or, I could get a new Llano board and APU, kick out my old Core 2 Extreme and power-sucking video card (hey, they were near-mandatory for Blu-ray watching when I bought 'em!) and get excellent HTPC performance in 1/3 the power usage, for the same price as just replacing the power supply.
 
Meh, ethics took its dying breath somewhere in the late 90s. Best to just accept it and know the line between legitimate business and scammers is gone. Capitalism at it's core is inherently evil without regulation. Oh, nevermind, that would take "big government", and we can't have that. Best to just let the hard working family christian companies self regulate.
 
Yeah, I trust the actual [H] tests much more than the labels, thanks so much for the benchmarks.
If efficiency comes with reliability, I see no reason for not choosing actual Gold PSUs over plain 80+.
And over the 3-5 years the computer or the reused PSU will be running, it makes a positive difference both environmentally and financially.
Even more so for intensive usage such as servers, folding, bitcoin mining, Firefox. ^-^

Paul, Kyle, last year's US average was 14 cents/kW if I remember searches I did a couple of months ago for another PSU comment. The DOE figure seems to be at the lowest end of the scale, I also pay far more in WA, not CA and some states have very high rates, not even talking about Hawaii. How much do you pay? 14 cents is also the standard price in France. Better use 15 cents to keep up with actual costs, and use yearly estimates instead of daily for meaningful comparison.

It can be far more efficient to run a 550w power supply that is 80Plus Bronze than, say a 1000w unit that is 80Plus Gold, just because someone may not be loading the unit up to the wattage level that is measured for the 80Plus certifications.
Actually, no, it is more efficient to use a 1000W Gold PSU even in that case.
The gain is 5% between Bronze and Gold PSUs of the same capacity for any identical system, but it is still 2-7% if you use a 550W Bronze PSU and a 1000W Gold PSU for the same systems (about 2% for a 300W system, up to 6-7% for a 440W system.)
 
Who can blame these PSU companies for doing this when the watchman is asleep at the switch?

I can, and I will.

It's not the fault of the police if a burglar breaks into my house when they are not around.

It's not even my fault if a burglar enters my house because I left the front door unlocked.

It's the fault of the burglar. He's the bad guy in all this, and guess what? If he enters my house, he's the one who is going to find himself bleeding out on my living room floor.

People and companies are responsible for doing the right ethical thing no matter what, even if no one ever would find out. This is the standard I hold myself to, and by extension I hold everyone else to as well.
 
I ignore the stickers and just buy Corsair, so far dozens of their psu's have never let me down.
 
Zarathustra[H];1037836792 said:
I can, and I will.

It's not the fault of the police if a burglar breaks into my house when they are not around.

It's not even my fault if a burglar enters my house because I left the front door unlocked.

It's the fault of the burglar. He's the bad guy in all this, and guess what? If he enters my house, he's the one who is going to find himself bleeding out on my living room floor.

People and companies are responsible for doing the right ethical thing no matter what, even if no one ever would find out. This is the standard I hold myself to, and by extension I hold everyone else to as well.

Haha, rhetorical question to setup the last page of the article ;)
 
<snip>

You are pretty brutal to supplies that do not meet other advertised recommendations, it seems unjust to be lax about the 80+ cert just because power in your area is apparently cheaper than pretty much anywhere else in the country.


Your thoughts are noted.
 
I don't think the certification came into play with the PSU I bought (Tt 750W Toughpower).

it was cause both [H] and Johnyguru had good things to say about it. At that point, I pretty
well felt like it was a good deal.

Its running all the stuff in the sig, and just plain does not have issues. Period.

There's a reason why review sites still exist, its because marketing is, well, marketing.
Its designed to sell products, not necessarily backup every claim made. Every mfg does
that in some form or another, we all have stories about that.

Regardless, hope it uncovers wool from eyes. Frankly any standardized test is only as good as
the data you put in, shit in, shit out as they say, and if you can game the system, people will do it.
 
While I commend your efforts with the article, I can't help but think you are preaching to choir.

I seriously doubt there's too many [H] readers that would buy a PSU from a brand that doesn't do consistantly well on your PSU testing period, regardless of what sticker they put on the box. (As evidenced by the comments here of people sticking to Corsair and Seasonic, for example) This is a niche community of very critical and bright minds after all.

If you want to really make a difference, rip out and test some PSUs from the top recommended Consumer Reports PCs (Dell Dimensions, HP Pavilions, etc.) and highly regarded brands (in the eyes of the general public) such as Alienware, Apple Mac Pro.

Then publically call them (the magazine recommending and the OEMs) out if and when they fail your usual testing standards. Hell for any poor parts for that matter. I was shocked to discover Apple using WD Blue drives for iMac, as my company has experienced a very high RMA rate with the whole WD Blue line.

I know you guys are already busy as hell and I commend the fact you cater to guys like me, but what I'd love even more is for a site like this to stand up and call BS on the millions of trash part PCs being sold to poor joe sixpack.
 
Thanks for the shout out!
Fold for the [H]orde!

We always tell people to get the most efficient PSU they can without breaking the bank. We push PSUs [H]arder than just about anyone around here, and that is saying something.

Often the major cost for folding if you are going all out (multi systems and OC-ing the heck out of them) isn't the cost of the system, but the cost of the power to run the system 24/7/365. A better PSU that can save you some power, with a modest price bump is well worth it in our circle.

 
I only buy power supplies that have passed rigorous testing. I plan on sticking to that rule and glad I can count on [H] to provide those results. Thanks! Great article and very well thought out.

Same here.

I also only buy power supplies that have both 80+ and Active PFC.

Almost all of my power supplies have been Antec. I have had a few that had caps die, but apart from having to RMA one, and replace caps in a couple others that were already out of warranty, they have been great.
 
Yeah, I trust the actual [H] tests much more than the labels, thanks so much for the benchmarks.
If efficiency comes with reliability, I see no reason for not choosing actual Gold PSUs over plain 80+.
And over the 3-5 years the computer or the reused PSU will be running, it makes a positive difference both environmentally and financially.
Even more so for intensive usage such as servers, folding, bitcoin mining, Firefox. ^-^

Paul, Kyle, last year's US average was 14 cents/kW if I remember searches I did a couple of months ago for another PSU comment. The DOE figure seems to be at the lowest end of the scale, I also pay far more in WA, not CA and some states have very high rates, not even talking about Hawaii. How much do you pay? 14 cents is also the standard price in France. Better use 15 cents to keep up with actual costs, and use yearly estimates instead of daily for meaningful comparison.


then move.. i live in washington as well and i pay way less than the 8 cents a KWh used in the article so go cry a river to your electricity company for ripping you off..
 
I don't think I ever bought a PSU based on the stickers, its like buying old video cards with them plastering "BLAAAAAAAAAAAAZING FAST 3D FASTER THEN EVAR!"
 
Perhaps in the future you should read all of the article ;) From page three:

I did read the entire article, and I've read most articles / reviews at [H] for some time, I'm aware of the methodology.

IMO The methodology should not excuse manufacturers from lying to customers. You are very strict on other advertised specifications. I guess I don't understand why this particular specification is okay to lie about and still receive significant praise. I have seen several PSU reviews in the past where supplies get recommends or awards when they do not meet their advertised 80+ spec, but I have not seen a recommend or award from a supply that did not meet it's advertised power output spec. This particular comment was not one of degree, but consistency. Why so tough on other specs, but soft on this one? MFRs are being encouraged to lie to consumers, and that's okay?

edit: started composing this during a conference call, before Kyle responded. Noting the noted, left for posterity.
 
I did read the entire article, and I've read most articles / reviews at [H] for some time, I'm aware of the methodology.

IMO The methodology should not excuse manufacturers from lying to customers. You are very strict on other advertised specifications. I guess I don't understand why this particular specification is okay to lie about and still receive significant praise. I have seen several PSU reviews in the past where supplies get recommends or awards when they do not meet their advertised 80+ spec, but I have not seen a recommend or award from a supply that did not meet it's advertised power output spec. This particular comment was not one of degree, but consistency. Why so tough on other specs, but soft on this one? MFRs are being encouraged to lie to consumers, and that's okay?

edit: started composing this during a conference call, before Kyle responded. Noting the noted, left for posterity.

Again as was stated:

While, in all honesty, we don&#8217;t care that much on a professional level about PSU efficiency as much as other factors when it comes to reviewing or selecting power supplies, manufacturers have decided that for bragging rights it will target users with this approach because it is something the company can use in its marketing. Since most of these PSU companies have found it so important, we hate to find that ECOS seems to have a significant difficulty with policing its certifications. We recently added our own version of 80 PLUS® certification testing to our PSU reviews and will continue to run ECOS' 80 PLUS® certification loads in our reviews. We started doing this because of the intense corporate focus on 80 PLUS® certifications, not because we find these all that useful. If a power supply comes in that misses 80 PLUS® certification numbers by a wide margin we will fail that unit.

Failing a unit is not exactly "soft".
 
Last edited:
then move.. i live in washington as well and i pay way less than the 8 cents a KWh used in the article so go cry a river to your electricity company for ripping you off..
Agree, let's trade places! ^-^ How much is your residential bill by KWH exactly? "Way less than 8 cents" seems unnatural.
I am with Puget Sound Energy and it's $0.10/KWH ($0.099435 to be precise), or rather $0.11/KWH for the whole bill.
The price per KWH is actually a sum of 6 individual rates, they don't want to make it easy on us.
There was actually an extra charge on top for "renewable energy credit" until April, so the total bill went cheaper by 1 cent per year or so. ^-^
I still think it's pretty cheap, not a rip-off.

Price per KWH (all rates add up to the same KWH usage figure):
  • Energy Charge: $0.085578
  • Electric Conservation Program Charge: $0.004336
  • Energy Exchange Credit: $0.006292
  • Federal Wind Power Credit: $0.000439
  • Merger Credit: $0.000325
  • Regulatory Asset Tracker: $0.002465

And there's a small Basic Charge and a City Tax of about $10 per month.
 
While the thought of companies gaming the system is quite deplorable, the article while correct on the points it touches on doesn't address non-80plus psu's. OK, so 80 plus is a sham, ignore it. So now HardOCP has suggested to consumers to go out and buy a PSU regardless of conversion efficency. While that is fine for most, we now have the strong case that someone will go out and get a 40% efficency PSU, and remember that a 40% psu takes roughly twice the draw to produce the same output as a 80% one (the rest is converted into heat).

So that being the case Paul Johnson, I'll be waiting for the next article where you tell people what they need to look for and why.
 
Personally I'll go for a high efficiency psu to help out the environment. I think we got to get out of the mentality of "its not going to save me any money so ill stick with the cheaper model". It may not save money but we will all be better off in the long run by consuming less energy, every little bit helps.
 
Interesting article there [H].

I look forward to the next sledge hammer of doom being taken to manufacturer's Power supplies that failed your tests and never did try to turn it about the right way.

Thank you again for being simply the best review site out there that is still free to read.
 
Ok, now apply this same article to energy star rating, epa vehicle milage rating (or even horse power/torque rating) or any "rating" scheme for that matter... Unless these rating companies do what [H] has been known to do, walk into a store, purchase a random sample and test it... Well manufacturers will always cheat because with out random off the shelf testing, they will never get caught.

On that note...ever notice that most epa milage ratings are estimated??? They don't even bother to test most of the time. Enjoy that 40 mpg eco car ;)
 
Agree, let's trade places! ^-^ How much is your residential bill by KWH exactly? "Way less than 8 cents" seems unnatural.
I am with Puget Sound Energy and it's $0.10/KWH ($0.099435 to be precise), or rather $0.11/KWH for the whole bill.
The price per KWH is actually a sum of 6 individual rates, they don't want to make it easy on us.
There was actually an extra charge on top for "renewable energy credit" until April, so the total bill went cheaper by 1 cent per year or so. ^-^
I still think it's pretty cheap, not a rip-off.

Price per KWH (all rates add up to the same KWH usage figure):
  • Energy Charge: $0.085578
  • Electric Conservation Program Charge: $0.004336
  • Energy Exchange Credit: $0.006292
  • Federal Wind Power Credit: $0.000439
  • Merger Credit: $0.000325
  • Regulatory Asset Tracker: $0.002465

And there's a small Basic Charge and a City Tax of about $10 per month.

Not saying that the article is off due to energy price, I still think all your conclusions are justified. Just chiming in on prices in Oregon. This was a breakdown of my bill for this month. Seems a bit higher than .08 with everything added in. It's lower for the power, but the distribution charge and Transmission charge put it up to over 11 cents an hr if I'm reading this correctly. Most of the electric in the NW "should be" some of the cheapest in the nation as we have (AFAIK) the largest hydro system in the US (13 dams and a 6 billion gallon reserve in canada we have a treaty for) and we were producing more than 16 million kwh per day due to high water flow and putting water through the dams (as fast as it would go without intorducing too much nitrogen and killing the fish) to prevent flooding. This summer/spring we were only using 6-7 mkwh a day in the greater Oregon areas due to mild temps (not as much AC usage). The DOE sells it's energy at cost and since we had an abundance of hydro and didn't need to get energy from the wind production farms, it was being sold to the energy companies and co-ops for dirt cheap. I work for the DOE so I know we can't sell power (by law) to the local companies for more than what it cost us to produce. The mark up is coming from PGE or whoever is selling it. :p

Basic Charge 9.00
Energy Use Charge
1000.000 kWh x 6.77800¢ 67.78
184.000 kWh x 7.50000¢ 13.80
Transmission Charge
1184.000 kWh x 0.23500¢ 2.78
Distribution Charge
1184.000 kWh x 3.11600¢ 36.89
 
Back
Top