View Full Version : Tiered PSU list
Proximon
11-03-2009, 06:29 PM
The most up to date version is being kept at newegg I think...
http://www.eggxpert.com/forums/thread/323050.aspx
But I don't know if I agree with that list completely... was wondering what the actual experts think.
I'm not a real expert, but I know enough to realize that lists like this are not sacred. They are just attempts by folks to make it simple for others. I appreciate the effort CompWiz has put into that list since he took it over.
I was thinking about rounding up all the actual modern adequately reviewed ;) PSU models and making a fresh start over at TH, something a bit more practical with a big "unknown" section and a big "probably trash" section. Emphasis would be on specific models instead of trying to assign values to model lines or brands.
A bit easier to do for someone like me, an advanced hobbyist, perhaps, rather than someone writing reviews.
Still, it's a big project and a commitment, so I thought I would get feedback. We spend a huge amount of time, collectively, at TH recommending builds to people. It's clear to me that it's not possible to give the average joe builder the education he needs on PSUs. I can and do link the guides here and at Hardware Secrets, but I'm sure most are too lazy to study those.
HOOfan_1
11-03-2009, 07:03 PM
That is just a rip off a list at Xstreme systems
Proximon
11-03-2009, 07:04 PM
I see David Hammock has this list that's very good, that I hadn't seen.
http://forums.extremeoverclocking.com/showthread.php?t=136602
Still, I would like something a bit less cluttered and more organized.
This list is better organized:
http://www.ocforums.com/showthread.php?t=589708
But there isn't enough of a ranking system.
dangman4ever
11-03-2009, 07:09 PM
Yeah that PSU list is full of PSUs in the wrong place without any justification. In addition, the criteria is a bit for each of the tiers is a bit fuzzy. What exactly is the difference between a Tier 2 and Tier 3 brand? In addition, as jonnyGURU pointed out years ago with many PSU lists: Not all PSUs within a certain brand are good. For example, in that list, the Xigmatek MC brand was listed in Tier 2. However, the MC 850W failed proper testing (http://hardocp.com/article/2008/10/16/xigmatek_nrpmc851_850w_power_supply/) while the MC 750 almost didn't pass testing. (http://hardocp.com/article/2009/03/18/xigmatek_nrpmc751_750w_power_supply/).
I do like your plan with a "unknown" and "probably trash" section with an emphasis on specific models rather than model lines or brands. However I do recommend that listed next to each recommended PSU the reason why it's recommended. I would also recommend a "good until proven otherwise" section since there are good PSUs out there that haven't been properly reviewed but are still recommended based on the performance of its actual manufacturer or a similar PSU model in the same line. Example: the EA430D and EA500D. The EA650 did pretty well under testing and is a Delta built PSU. As such, the lower wattage EA430D and EA500D should be just as good as the higher wattage EA650 as they are based on the same platform.
So basically: Make sure that you define your exact criteria for each tier (like difference in amperage, ripple results, etc) and try to justify to each and every PSU you recommend, whether through a proper test link (HardOCP.com, JonnyGURU, PCPerspective, etc) or a line of logical reasoning. In fact, sorting each tier by amperage like Davidhammock did sounds like a good idea as well. Then again, the "single rail" recommendation in that list is a bit wrong as well.
HOOfan_1
11-03-2009, 07:10 PM
problem is that people will still ask for suggestions and just ignore the list.
Or if they read the list, they won't know how much power they need and will ask anyway.
Proximon
11-03-2009, 07:10 PM
That is just a rip off a list at Xstreme systems
I thought that one was out of date though?
If this is the list, it hasn't been touched in 18 months
http://www.xtremesystems.org/forums/showthread.php?t=108088
HOOfan_1
11-03-2009, 07:16 PM
I thought that one was out of date though?
If this is the list, it hasn't been touched in 18 months
http://www.xtremesystems.org/forums/showthread.php?t=108088
Just compare the lists...obviously a rip off...they simply plugged in some new PSUs where they believe they go...and the list is pretty flawed
dangman4ever
11-03-2009, 07:18 PM
I thought that one was out of date though?
If this is the list, it hasn't been touched in 18 months
http://www.xtremesystems.org/forums/showthread.php?t=108088
It is out of date but it was still ripped off and modified from that Xtremesystem link. The reasons for each tier, many of the PSU brands listed, etc in the Newegg link is the exact same as the Xstreamsystem albeit with more additions.
It doesn't look like they actually moved any of the original PSU brands into what should be their rightful tier. FFS, they list the Truepower II as a tier 3 PSU with the criteria of "High Quality and Stability" despite the fact that the Truepower II PSUs had a high death rate and were not of good quality due to the following reason: Antec/CWT specified/used caps (Fuhjyyu caps) that did not respond well to heat. Unfortunately, Antec made the decision to have the PSU fan only ramp at really high temps in an effort to keep the PSU quiet. However, just medium to high temps were enough to screw up the caps so really high temps were murderous. Over time, this degraded the quality and performance of those PSU series.
So yeah, as HOOfan_1 said: the list is pretty flawed. So if you do create a new list, do not base ANY brand or specific model recommendation on the Newegg or Xstreamsystem lists. Just start fresh as you said earlier.
Proximon
11-03-2009, 07:26 PM
Thanks for the feedback! Separate sections by 12V amps for all the adequate tiers, and then the unknowns and such could be listed by advertised watts. I like that.
Yes tiers have to be defined better if it's done that way, but I wouldn't presume to know better than you experts. I would probably just have a date and link to the review, such as
Tier 2
- Corsair 950TX 10/17/09 jg (http://www.jonnyguru.com/modules.php?name=NDReviews&op=Story&reid=170)
Again, just considering all this and getting input. Not committing to anything or such as yet.
Edit: I realized it would be politer to link to the first page of any review.
370forlife
11-03-2009, 08:10 PM
I doubt the guy at eggxpert has any idea what a power supply is. He firmly believes what he copied up there and occasionally updates is the way it is, no if's, an's, or buts...
I like how he puts delta and AcBel built psu's in tier 1 and 2, while he puts delta and AcBel in tier 4 and 5. He also thinks very, very highly of OCZ power supplies and recommends them all the time. I don't think he understands how OEM'ing works
Proximon
11-17-2009, 04:29 AM
So I've made a start with a rough draft of the intro etc. Just getting a feel for how it might work out and exactly how much work it will be.
Does anyone have any ideas on how the 12V rail amp categories would be broken down?
To start I went with 65-75A, which includes most of the better 850W units and also the TPQ-1000... that's as far as I've gone.
50-64?
40-49?
Does anyone have a feel for what might be best there?
Proximon
11-17-2009, 06:19 PM
OK so this is all very rough. I'm going to post this here so that you guys can really tear into it if you like... consider a somewhat evolved idea at this point, not even a rough draft.
I'm just getting a feel here for the work involved and how it might go, Still not committing to anything.
---------------------
Ranked Power Supplies
The purpose of this list is to provide good information that will help anyone decide what PSU to purchase.
For many years there has been a Tiered PSU List that gets used to place various brands and model lines into categories. This has helped the community a great deal but it does have some problems. There have also been some attempts at updating this, but they have not been successful.
Desktop computer power supplies need to be judged by the following criteria:
1. Performance under stress. High operating temperatures combined with heavy loads can bring many supposedly good power supplies to their knees. Most so-called reviewers fail to properly stress a PSU during testing.
2. Longevity. This is not an easy thing to judge. Sometimes the historical performance of a particular OEM or design can be used, but mostly expert evaluation of the various internal parts and assembly is the only way to predict how long a given unit will last.
Beyond these two are more specific criteria, but these may change in importance depending on the user. I list here my personal order and yours may be different. However, all of these items are important, and none of them should be ignored.
3. Warranty. Not only duration but actual service quality.
4. Voltage Regulation. The ability of the unit to supply stable voltage under different demands, from low to high.
5. Voltage Ripple. The amount of fluctuation remaining in the DC current after conversion from AC. Can only be seen and measured on an oscilloscope or such.
6. Noise levels. The plural is there because this can change depending on load and ambient temp. Usually this is only the fan noise, but even passively cooled units can generate some noise.
7. Cabling. Cable length, modularity, sleeving, cable type.
8. Aesthetics. Color, fan lighting, etc.
In modern desktop computers, especially those used at least partly for graphically intense games, total wattage is not as important as maximum amperage on the 12V rail(s). For this reason, this list is broken down not by wattage, but by 12V amps.
The ranks will be as follows in each amperage range:
Rank 1 - Will contain a maximum of 4 units. If another is added that causes 5 in this rank, one must be removed to rank 2. As power supplies become increasingly sophisticated and of higher and higher quality, the contents of these ranks must also shift.
Rank 2 - Will contain former rank 1 PSUs plus any current models that compete well with them. This list will be large, but somewhat self-limiting as older models go out of production.
Rank 3 - Will contain recommended PSUs that can't quite compete with the better ones but may be desireable for budget or availability considerations.
Rank 4 - Will contain barely adequate PSUs that should only be selected in desperate circumstances.
Rank 5 - Will have units with serious flaws in design, parts, or performance. Not recommended for any desktop. Any failure to meet current ATX specs will land a unit here.
Unknown - This will contain units that cannot yet be placed in any other category. Someone has looked at them and there is some doubt about quality, but nothing concrete is known.
Unlisted - PSUs that no one has bothered to look at or comment on will NOT appear in this list.
Key
jg = Jonnyguru
hs = Hardware Secrets
H = [H]OCP
O3 = Overclock 3D
AT = Anandtech
TH = Tom's Hardware
SP = Silent PC Review
XB = XBit Labs
italics = No review but quality inferred from OEM, brand, and model line
Example:
65-75 Amps
Rank 1
Corsair HX850 70A jg (http://www.jonnyguru.com/modules.php?name=NDReviews&op=Story&reid=153)
Antec SG-850 65A jg (http://www.jonnyguru.com/modules.php?name=NDReviews&op=Story&reid=101)
Seasonic M12D 850W 70A H (http://enthusiast.hardocp.com/article/2009/01/14/seasonic_m12d850_power_supply/)
XFX 850W Black 70A HS (http://www.hardwaresecrets.com/article/808)
Rank 2
Corsair 850TX 70A jg (http://www.jonnyguru.com/modules.php?name=NDReviews&op=Story&reid=126)
PC P&C Silencer 910W 74A H (http://enthusiast.hardocp.com/article/2009/10/22/pc_power_cooling_silencer_910_psu_review/1)
ENERMAX REVOLUTION85+ ERV850EWT 70A jg (http://www.jonnyguru.com/modules.php?name=NDReviews&op=Story&reid=150)
Antec TPQ-1000 70A jg (http://www.jonnyguru.com/modules.php?name=NDReviews&op=Story&reid=170)
OCZ Z850M 71A H (http://enthusiast.hardocp.com/article/2009/10/08/ocz_z_series_z850m_850w_power_supply_review/1)
Thermaltake Toughpower XT 850 H (http://enthusiast.hardocp.com/article/2009/09/21/thermaltake_toughpower_xt_850w_psu_review_update/)
Hondacity
11-18-2009, 02:09 PM
i wouldn't get that seasonic m12d or that xfx850....
Proximon
11-18-2009, 04:02 PM
I've been working on this over here as well:
http://www.tomshardware.com/forum/270297-28-ranked-list
I've added 80 Plus color coding and looking at some other things.
I'm open to moving anything around. This is an example of how it might look, not meant to be definitive at this point.
dangman4ever
11-18-2009, 07:29 PM
i wouldn't get that seasonic m12d or that xfx850....
Umm, and why is that?
Hondacity
11-18-2009, 07:40 PM
Umm, and why is that?
id rather go with the antec850 and corsairs:D
HOOfan_1
11-18-2009, 10:35 PM
Although it is not an exact science trying to actually produce a tiered list of PSUs, I would say that the PC P&C Silencer 910 and the Enermax Revolution 85+ both belong in the same ranking as the M12D, HX850, SG-850, and XFX Black Edition
The Silencer is the same platform as the M12D and XFX afterall, and certainly performs as well. The Revolution is certainly up there too.
Also you might want to consider breaking down the list into modular and non-modular and perhaps add further description.
Proximon
11-18-2009, 11:44 PM
Once we get all the format straightened out we'll get to the details, and it looks like I'll be getting some help.
I wanted to limit the number of rank 1's, but if you think it should be larger I might do that, or I'm certainly willing to listen to a different list of top 4.
We have some changes made already on the other forum, but I can't easily cross-post because the BBCode is different.
We'll very much need opinions from this forum to pull this off though!
Zero82z
11-19-2009, 10:50 AM
The entire concept of a tiered list is just silly in my opinion. It's better to rank PSUs relative to each other than waste time trying to make them fit into ranks.
jonnyGURU
11-19-2009, 12:43 PM
The entire concept of a tiered list is just silly in my opinion. It's better to rank PSUs relative to each other than waste time trying to make them fit into ranks.
QFT
C'DaleRider
11-19-2009, 01:29 PM
The entire concept of a tiered list is just silly in my opinion. It's better to rank PSUs relative to each other than waste time trying to make them fit into ranks.
Like JG said, truer words are rarely spoken.
Absolutely no sense in doing a Tier list....most are wholly incorrect or just reflect the list maker's biases in companies.
Attempting a Tier list is, at least to me, nothing more than tilting at windmills.
Proximon
11-19-2009, 04:33 PM
I understand and respect those opinions, but the problem is that people use them... a lot.
People who know about power supplies can easily choose which one is best, but those who don't know need some kind of cheat sheet. The original list has 209,000 hits, the eggspert list has 58K.
I don't think it's easy to get up to speed on power supplies. I spent probably between 10 and 20 hours studying the stickies here and various other resources before I had any sort of clue at all about them.
How many at home builders are really going to take the time to do what essentially amounts to a 101 level college class to get an education on PSUs? 3-5%?
Even if I only save a few thousand people from buying junk PSUs, it will be worth it.
Redbeard
11-19-2009, 05:05 PM
A lot of marketing is done in a "Good - Better - Best" tiered structure, so any more than 3 tiers might be asking for trouble.
As far as what I think would be useful to consumers, I have my own opinions (which are not those of my employer).
GOOD
-NEVER outside of ATX performance specifications
-80 Plus efficiency
BETTER
-NEVER outside of ATX performance specifications
-80 Plus efficiency or better
-At least 1 outstanding feature (80 Plus Bronze or Silver, exceptional VREG, exceptional ripple, etc)
BEST
-NEVER outside of ATX performance specifications
-80 Plus efficiency, Bronze or higher
-At least 3 outstanding features (example: 80 Plus Bronze, +12V ripple under 50% of ATX spec, VREG +/- 3%, longer than average cables for full tower cases, super-quiet fan, outstanding value for the money, etc)
Also, I think it should be clearly marked which products are modular and which are not.
This is the type of guide I think consumers would like.
EDIT - while I agree that a tiered ranking list is vague and general and wouldn't be as good as focusing on each specific part's reviews, I think the purpose of this guide is to give somebody with limited shopping options a level of comfort with the products they're offered. A "cheat sheet" that labels reliable parts. I mean, if you can't find your first choice in PSU, knowing which other ones are decent would be very helpful.
Zero82z
11-19-2009, 05:27 PM
I understand and respect those opinions, but the problem is that people use them... a lot.
People who know about power supplies can easily choose which one is best, but those who don't know need some kind of cheat sheet. The original list has 209,000 hits, the eggspert list has 58K.
I don't think it's easy to get up to speed on power supplies. I spent probably between 10 and 20 hours studying the stickies here and various other resources before I had any sort of clue at all about them.
How many at home builders are really going to take the time to do what essentially amounts to a 101 level college class to get an education on PSUs? 3-5%?
Even if I only save a few thousand people from buying junk PSUs, it will be worth it.
Then why not just make a list of good PSUs instead? That way people will know what PSUs to look for without having to wade through all the nonsense of a tier-based ranking system.
Proximon
11-20-2009, 02:41 AM
Redbeard, I have had several votes for modularity now, so I see I'll have to include it.
Also, I like the way you have managed to quantify a ranking system... We had pretty much decided that 80 Plus would be required for any recommended PSU... I just thought there might be a few older units that would need to be allowed in. I have only seen one so far, but I may have a few more before we're done.
I should probably explain myself a bit more. I maintain a stickied guide over at TH and we, collectively, field maybe an average of 30-40 requests for build advice per day. As far as I know it's the busiest such forum on the net. No one can get to every post every day of course. While many people that help out there do understand enough about PSUs, many more just jump in to lend a hand without knowing.
I already have a few knowledgeable volunteers that want to help with this project and expect to get more, without even asking. It's a project that seems to have caught on quickly with the people that hand out advice regularly.
Having this list takes it out of the realm of opinion. This list will help us out, if in no other way it will bring everyone handing out advice at TH onto the same page. It establishes a consensus, and I'm hoping the many real experts on this board will let us know if they feel there is serious error about how we place a PSU... I promise there won't be many of those.
Zero82z, without some ranking that OCZ 550 F4tali1y (or however it's written) is going to look just as good or better than a 550VX or TP-550. Well, I don't think anyone knows for sure about that particular OCZ unit, but you get the idea.
Sorry I'm writing a wall of text here, but one last thing:
You might be surprised by how many PMs and forum requests I have to deal with from people in countries without access to large incomes or big selections. I also see some of them coming here. These people need some idea of what is junk and what might be OK for them. They are why I keep hanging on to that rank 4 idea.
Zero82z
11-20-2009, 11:41 AM
Zero82z, without some ranking that OCZ 550 F4tali1y (or however it's written) is going to look just as good or better than a 550VX or TP-550. Well, I don't think anyone knows for sure about that particular OCZ unit, but you get the idea.
I didn't say you shouldn't rank them. I just said that you shouldn't use a tiered ranking system.
Redbeard
11-20-2009, 02:51 PM
I didn't say you shouldn't rank them. I just said that you shouldn't use a tiered ranking system.
Isn't that the same thing?
Ranking units on a 1-10 scale is tiered, too.
Proximon
11-20-2009, 03:12 PM
Ranking units on a 1-10 scale is tiered, too.
Yes, exactly. Some sites use 1-10, some percentage, some do a pass/fail/recommend... it's all ranking.
to put my ranks more simply, they are
1. the best
2. recommended
3. acceptable
4. only if you have to
5. never
And then I have a spot for anything I can't fit into one of those.
Redbeard
11-20-2009, 04:26 PM
Yes, exactly. Some sites use 1-10, some percentage, some do a pass/fail/recommend... it's all ranking.
to put my ranks more simply, they are
1. the best
2. recommended
3. acceptable
4. only if you have to
5. never
And then I have a spot for anything I can't fit into one of those.
I think you might be making it too complex.
I would stick with
1 - The Best
2 - Recommended
3 - Acceptable
Then if you absolutely have to, keep a "NOT RECOMMENDED" list, and put parts from your 4th or 5th tier down there.
I think the real benefit of this has to be for somebody who is just overwhelmed by a REAL review of the product and just wants a cheat sheet. As such, keeping it simple is necessary. Nobody needs 5 separate tiers, even 3 is pushing it. HardOCP has two tiers, really: PASS and FAIL, and occasionally exceptional products get Gold or Silver awards.
But then again, this is all my opinion.
Proximon
11-20-2009, 08:33 PM
Simple is good, it's a big project.
Perhaps the tier 4 is too much a personal desire rather than a desirable feature. I'll run it by the guys on the other forum.
RagingDragon
11-22-2009, 03:15 PM
I like Rebeard's suggestion of best, recommened, and acceptable. Just leave the junk units off the list entirely - that'll keep the list smaller and easier to understant, it'll also keep you out of trouble with brands/manufacturers upset about being put on the "bad" list. Anyway since the worst junk is sold under numerous obscure and short lived brands I doubt it's feasible to maintain an compreshensive "bad" list.
Proximon
11-23-2009, 02:13 AM
Since it's now very much a group project, we'll see how this evolves.
It was not so long ago that I knew nothing at all about power supplies and I recall well the first time I saw the original tiered PSU list. It drove home that a recognizable name brand does not mean a quality part.
When I see a list of good PSUs, I have no idea whether it's a personal list of favorites or a well researched effort. When I see a list that includes a broad range of units of all kinds and several levels of quality, I'm much more likely to give it credence.
Proximon
01-14-2010, 05:08 AM
I thought I would give an update. This project is under way now. We are using an online collaborative database called Dabble.
You can view our progress at any time here:
Ranked PSU List (https://rankedpsulist.dabbledb.com/page/rankedpsulist/aninOXDN#)
I have several volunteers and in just a couple days we have a good start made. Lots of work left to do, for sure.
Hondacity
01-14-2010, 08:50 AM
thats like 20% complete...
my 2cents
Spectre
01-14-2010, 10:03 AM
The XFX 650W is Seasonic, via Seasonic. PC Power had nothing to do with it.
Nexus RX-850 is ATNG not Hipower.
OCZ Z series are Sirfa.
What does the rank mean?
Redbeard
01-14-2010, 11:51 AM
I thought I would give an update. This project is under way now. We are using an online collaborative database called Dabble.
You can view our progress at any time here:
Ranked PSU List (https://rankedpsulist.dabbledb.com/page/rankedpsulist/aninOXDN#)
I have several volunteers and in just a couple days we have a good start made. Lots of work left to do, for sure.
What do the numbers on the top left mean, and what do the ranks correspond to?
Perhaps a legend or key would be useful?
Proximon
01-14-2010, 01:49 PM
Thanks for the corrections Spectre, I'll adjust those... I guess the "PC P&C" printed on the circuit board of the XFX had me confused.
This is just the database we are working on. When we get closer to completion we'll integrate it into a web page or thread with a key, etc. 20% is probably pretty optimistic... more like 10-15% done. Notice that about half of what is there is the easier 1000W+ range. Well, in all fairness the 850W area is pretty easy too. Everybody makes "500W" units, and that's going to take a while to sort through.
The numbers at the top left are 12V amps, as per the power distribution on the label. They are ranges, pretty much centered on what is normal for that wattage. It's sorted that way because it's a little bit more practical than total watts.
So, for instance, most decent 850W PSUs will have a 70A 12V rail... so this is considered the norm, and a range is created around that. 65-75 in this case. We currently have anything above 75A grouped together.
A lot more thoughts on our thread at TH:
http://www.tomshardware.com/forum/270297-28-ranked-list
HOOfan_1
01-14-2010, 10:21 PM
Thanks for the corrections Spectre, I'll adjust those... I guess the "PC P&C" printed on the circuit board of the XFX had me confused.
PC P&C doesn't make PSUs...Seasonic makes their Silencer line and apparently they are using the circuit boards they had silk screened for PC P&C. PC P&C likely left them with a ton when they pulled their products a couple of months ago.
forsunny
01-31-2010, 02:15 PM
I feel that efficiency should be given a much higher weightage in judging quality of the power supplies.
Zero82z
02-01-2010, 01:59 PM
I feel that efficiency should be given a much higher weightage in judging quality of the power supplies.
Why do you say that? The quality of power supplied by a PSU is generally of much greater importance than its efficiency.
370forlife
02-01-2010, 03:35 PM
I feel that efficiency should be given a much higher weightage in judging quality of the power supplies.
Yes, as some of the units in the bargain basement roundup have better efficiency than server grade units...makes sense....
ef3393
02-01-2010, 07:35 PM
I think so especially good efficiency will reduce heat and improve reliability!!
10% efficiency make huge difference, for example 80% vs 90% efficiency
80% efficiency=100Wac input, 80Wdc output, 20W turn into the heat
90% efficiency=100Wac input, 90Wdc output, 10W turn into the heat
the 10% efficiency give you 12.5% more output and cut the heat to half
the heat also related to fan noise too!!
________
Coach Handbags (http://icoachhandbags.com/)
Zero82z
02-01-2010, 07:39 PM
I think so especially good efficiency will reduce heat and improve reliability!!
10% efficiency make huge difference, for example 80% vs 90% efficiency
80% efficiency=100Wac input, 80Wdc output, 20W turn into the heat
90% efficiency=100Wac input, 90Wdc output, 10W turn into the heat
the 10% efficiency give you 12.5% more output and cut the heat to half
the heat also related to fan noise too!!
Efficiency isn't everything: http://www.hardocp.com/article/2010/01/22/abs_majesty_mj1100m_1100w_power_supply_review/8
ef3393
02-01-2010, 08:19 PM
I understand efficiency is not everything, only when other electronical performance is the smae then better efficiency is very important.
________
Penny stocks to watch (http://pennystockpicks.net/)
ef3393
02-01-2010, 08:25 PM
Thus 80Plus Gold MJ1100-M is really a poor PSU from my point of viewdue to the poor output quality, and I admit a porr PSU with 80Plus Gold is still nothing.
However to compare a group of top performance PSU the efficiency will become veru important.
________
HURT FROM ZOLOFT (http://www.classactionsettlements.org/lawsuit/zoloft/)
Plimogz
02-01-2010, 10:36 PM
I understand efficiency is not everything, only when other electronical performance is the smae then better efficiency is very important.
QFT.
Let me just expand on that to say that IMO as long as the PSUs you are comparing are within ATX spec, efficiency could be a prime concern. In cases where you are "system thermal-envelope constrained" for instance. Or perhaps if you are very concerned with system power usage (e.g. You are deploying several dozen 24/7 machines). I suppose effeciency is also pretty tighly coupled with noise-output, so quiet computing buffs might also use higher 80+ certifications as a good indication that a PSU will operate quietly even under heavy loads.
On the other hand, from an overclocker's POV, I know that effeciency takes a distant back-seat to voltage stability and low ripple.
Loving the forums, BTW!
Proximon
02-01-2010, 10:57 PM
I'm willing to play along :)
So, you are saying that the OCZ Z850M reviewed here:
http://enthusiast.hardocp.com/article/2009/10/08/ocz_z_series_z850m_850w_power_supply_review/1
Which is 80Plus Gold
Should be rated as highly as the Antec SG-850 reviewed here:
http://www.jonnyguru.com/modules.php?name=NDReviews&op=Story&reid=101
Which is 80Plus Bronze?
If not, then perhaps you could give a specific instance?
Our goal is to have only 80Plus certified units in tiers 1-3, however there will be a few units initially that are still available and never got certified. OCZ700SXS and Xclio Diamondpower 880W come to mind.
Plimogz
02-02-2010, 12:06 AM
Why you wouldn't link to two OklahomaWolf reviews is beyond me:
OCZ Z series 850W (http://www.jonnyguru.com/modules.php?name=NDReviews&op=Story&reid=167)
Antec Signature 850W (http://www.jonnyguru.com/modules.php?name=NDReviews&op=Story&reid=101)
So, Proximon, I assume I'm not the one you're playing with right now, but I'll go ahead and butt into your game anyway :)
Comparing both Cold and Hot load tests, the Signature maintains voltages closer to nominal across all rails, across all data points. Comparing noise/ripple is a joke, I've got both reviews opened side-by-side here: The Antec mops the floor with the OCZ. The efficiency advantage however is squarely in OCZ's court. (i.e the Z-series' avg. efficiency comes in around 90% whereas the Signature's numbers put it at about 85%)
So the only conceivable reason to rate the OCZ as high (or perhaps even higher) than the Antec would be on grounds of efficiency.
I'm willing to play along :)
So, you are saying that the OCZ Z850M [...] Which is 80Plus Gold
Should be rated as highly as the Antec SG-850 [...] Which is 80Plus Bronze?
If not, then perhaps you could give a specific instance?
Now, if a prospective buyer is very interested in saving that 5% energy (and remember that in some cases people are using Air Conditioning to keep room temperatures in check, so any efficiency gain at the PSU is nearly doubled through AC energy savings) it may be worthwhile for such a customer to go with the more efficient gold-rated unit.
Not me though, I'm all about the stable, ripple-free rails. So any rating system which would rate the Z850 higher overall than the SG-850 would be highly suspect in my eyes. I think that 80+ ratings should absolutely be shown, and certainly used as a quantifiable means to break potential ties between models, but nothing more.
Proximon
02-02-2010, 01:52 AM
Why you wouldn't link to two OklahomaWolf reviews is beyond me:
Habit. I do a lot of linking and prefer to send some hits to the other quality reviewers. They all need encouragement ;)
ef3393
02-02-2010, 02:35 AM
still thinking the Z-850 is totally not in the same class with SIG-850 then even the efficiency is much higher could not help, there are more and more 80Plus Gold units such as Modu 87+ and X have better output quality than Z-850
________
Classaction Settlements (http://classactionsettlements.org/)
C'DaleRider
02-02-2010, 05:40 AM
In my evaluation of a power supply to purchase, efficiency, while important, falls behind construction quality and electrical output quality (noise, ripple, voltage regulation.) Ultra high efficiency does not equate to construction quality nor quality of electrical output.
Concerning the Z-Series vs. the Antec Sig. line, I'd buy the Antecs and forget the OCZ's every day of the week. True, the OCZ's are more efficient, but that's all they have on the Antecs. In every other meaningful metric, the Antecs are simply better than the OCZ units....as is the Seasonic X's and the Corsair HX.
Proximon
02-02-2010, 04:16 PM
Perhaps when forsunny comes back we'll get an example of what is meant.
vBulletin® v3.8.1, Copyright ©2000-2012, Jelsoft Enterprises Ltd.