[OLD] PSU Tier List 3.0 (Legacy)

[Zando_]

So.... not a technical PSU nerd but uh... Why are the RMi/RMx/RM/HX PSUs on the same level as the AXi? And why is the T2 lower than those, and why is the Focus Plus Gold which is known to have issues with some GPUs on the same tier as the EVGA T2 and P2? And why are the G/G1+/GQ PSUs on the same level as the G2, and why is that also on the same level as the T2 which is more of an AXi competitor? Is there a method to the madness or are you kinda just sorting stuff at random? And why the heck is there a Silver PSU on the same level as the AX and some RM PSUs? 

 

Your A+ and A lists seem really confused as to what they are. 

[flibberdipper]

2 minutes ago, Zando Bob said:

Your A+ and A lists seem really confused as to what they are. 

The whole list seems highly confused as to what it's supposed to be.

 

^{Kinda like it's just keyboard tested.}

[hello_there_123]

5 minutes ago, Zando Bob said:

 Why are the RMi/RMx/RM/HX PSUs on the same level as the AXi?

Because we can't be bothered enough to create a tier S with only a few highest end PSUs. 

 

6 minutes ago, Zando Bob said:

 And why is the T2 lower than those, and why is the Focus Plus Gold which is known to have issues with some GPUs on the same tier as the EVGA T2 and P2?

First question: s i n g l e r a i l 
Second question: That's what the limitations of a tier list do. Unless you do it in spreadsheet form where you can rank PSUs individually within a tier...

 

7 minutes ago, Zando Bob said:

And why are the G/G1+/GQ PSUs on the same level as the G2 

Good question. I've been asking the other tier list creators this for a long time now...

 

7 minutes ago, Zando Bob said:

and why is that also on the same level as the T2 which is more of an AXi competitor?. 

Tier list limitations, unless in spreadsheet form. Unfortunate, but that's what happens ig. 

 

8 minutes ago, Zando Bob said:

And why the heck is there a Silver PSU on the same level as the AX and some RM PSUs? 

Again, limitations of the tier list. I proposed a tier B+ for added flexibility, but no one else apart from Luke supported it...

[Damascus]

Where does this guy sit?

https://www.fspgroupusa.com/ecommerce/ipc-psu/flex/500w/fsp500-50fspt.html

7 minutes ago, hello_there_123 said:

Because we can't be bothered enough to create a tier S with only a few highest end PSUs. 

So... laziness 

7 minutes ago, hello_there_123 said:

First question: s i n g l e r a i l 
Second question: That's what the limitations of a tier list do. Unless you do it in spreadsheet form where you can rank PSUs individually within a tier...

Most good tier lists do that, look at all of the respectable motherboard ones 

7 minutes ago, hello_there_123 said:

Good question. I've been asking the other tier list creators this for a long time now...

 

Tier list limitations, unless in spreadsheet form. Unfortunate, but that's what happens ig. 

 

Again, limitations of the tier list. I proposed a tier B+ for added flexibility, but no one else apart from Luke supported it...

More laziness and internal strife...  What is going on here guys?  If you want this list to be an authority, put in the effort to make everything clear and concise instead of using a barely consistent grade school level ABCD scoring system.  (Not trying to pick on you specifically, everyone involved needs to  either get this into a well organized list or stop cramming it down the collective tech communities throats)

[LukeSavenije]

46 minutes ago, Zando Bob said:

Why are the RMi/RMx/RM/HX PSUs on the same level as the axi?

because list limitations, on these units alone we could make 5 different tiers, but this would get way too cluttered

 

47 minutes ago, Zando Bob said:

why is the T2 lower than those,

read the notes

 

47 minutes ago, Zando Bob said:

why is the Focus Plus Gold which is known to have issues with some GPUs

will be moved in 1.5

 

48 minutes ago, Zando Bob said:

Is there a method to the madness

yes, read the notes

 

2 hours ago, hollyh88 said:

so as far as i can see.. any focus plus bought after 2018 havent seen issues?

it has actually shown worse issues, with reports from pceva that it has 268mV of ripple at 120%, which is more than double that atx spec allows it to be

 

41 minutes ago, flibberdipper said:

The whole list seems highly confused as to what it's supposed to be.

i mean, show me the perfect method, and I'll listen

 

39 minutes ago, hello_there_123 said:

proposed a tier B+ for added flexibility, but no one else apart from Luke supported it...

actually lienus seems to support it as well

 

24 minutes ago, Damascus said:

So... laziness

no, stop of clutterness

 

25 minutes ago, Damascus said:

Most good tier lists do that, look at all of the respectable motherboard ones 

I've seen those... or rather, i made one (or two, depends on who you ask)

 

so tldr: we can make a list for you that has 300 PSUs in it, all in their exact level, and we would have 300 tiers

[flibberdipper]

2 minutes ago, LukeSavenije said:

i mean, show me the perfect method, and I'll listen

Well you see, instead of regurgitating on-paper bullshit and just going "I like this one more so it goes here" you buy the equipment, get the power supplies, and actually put them through the ringer.

 

Shits the same as ranking an Audi A5 over the Maybach S550 because "I like the styling more lol" when the S550 is the superior car in every aspect.

[Damascus]

Still curious about mine ( ͡° ͜ʖ ͡°)

 

https://www.fspgroupusa.com/ecommerce/ipc-psu/flex/500w/fsp500-50fspt.html

 

5 minutes ago, LukeSavenije said:

no, stop of clutterness

The more granular a ranking system, the more value it has.  Why do you think your report card changed from ABCD to x/100 between grade school and high school/university?  It's because the letter system is an objectively inferior product designed for children 

5 minutes ago, LukeSavenije said:

I've seen those... or rather, i made one (or two, depends on who you ask)

 

so tldr: we can make a list for you that has 300 PSUs in it, all in their exact level, and we would have 300 tiers

And you would have a superior, more reliable product.  That would allow you to use, among other things, top and bottom percentile, wherein you recomend everything above "X" in the ranking, as opposed to an amorphous letter or broad number.  

[LukeSavenije]

Just now, Damascus said:

Still curious about mine ( ͡° ͜ʖ ͡°)

I'll look into it, reports later

 

8 minutes ago, flibberdipper said:

buy the equipment, get the power supplies, and actually put them through the ringer.

sorry, i don't have 5000 dollars spare for a load tester, a soldering iron, a oscilloscope, some PSUs, a designer so i can actually make money off of this and i don't think cm would let a 17 year old test a psu on their testing department

 

10 minutes ago, flibberdipper said:

I like this one more so it goes here

then tell we which shouldn't deserve their tiers, I'll throw it in the group

[hollyh88]

19 minutes ago, LukeSavenije said:

it has actually shown worse issues, with reports from pceva that it has 268mV of ripple at 120%, which is more than double that atx spec allows it to be

 

okay but those issues only have been created once you come close to the max output of the wattage right? for the focus plus gold 850watt an wattage of around 600 should easily be done without issues right? or not? can you explain a bit more about that

 

"so as far as i can see.. any focus plus bought after 2018 havent seen issues? i just bought an focus plus 850 watt psu. and as for the 850 watt why so much.. well ehh the 750 was more expansive. so i said fuck it and bought a 850 watt

i will use nvidia cards so thankfully i have no issues with wattages spiking like a crazy thing "

something like this with lets say my 2600x at 4.1 and now gtx 1070 sc from evga but later probably the msi gaming trio x 2070 super wouldnt create issues for it right? 

[flibberdipper]

14 minutes ago, LukeSavenije said:

sorry, i don't have 5000 dollars spare for a load tester, a soldering iron, a oscilloscope, some PSUs, a designer so i can actually make money off of this and i don't think cm would let a 17 year old test a psu on their testing department

So then don't spew hot garbage when you're clearly not capable of doing a tier list correctly. Not that hard.

[hello_there_123]

16 minutes ago, hollyh88 said:

okay but those issues only have been created once you come close to the max output of the wattage right? for the focus plus gold 850watt an wattage of around 600 should easily be done without issues right? or not? can you explain a bit more about that

Yes. Specifically 120% of the rated wattage output. 

 

17 minutes ago, hollyh88 said:

something like this with lets say my 2600x at 4.1 and now gtx 1070 sc from evga but later probably the msi gaming trio x 2070 super wouldnt create issues for it right? 

Correct.

[Zando_]

1 hour ago, hello_there_123 said:

Because we can't be bothered enough to create a tier S with only a few highest end PSUs. 

Seems like that'd make more sense, isn't there a risk rn of someone buying an RMi and thinking they're getting the same level of PSU as an AXi? 

1 hour ago, hello_there_123 said:

First question: s i n g l e r a i l 
Second question: That's what the limitations of a tier list do. Unless you do it in spreadsheet form where you can rank PSUs individually within a tier...

(can I get a quick explanation of the single rail vs multiple rail thing that actually makes sense instead of being a load of clumsily worded technical jargon designed to just confuse the person reading it instead of informing them? What I just described is all I've ever gotten from the self proclaimed PSU guys here)

Spreadsheet would be solid, I've seen peeps do that with mobo VRMs for the Zen 2 chips and it's a super handy list. Could just add a couple more tiers though to split stuff out a bit more logically.

1 hour ago, hello_there_123 said:

Good question. I've been asking the other tier list creators this for a long time now...

 

Tier list limitations, unless in spreadsheet form. Unfortunate, but that's what happens ig. 

 

Again, limitations of the tier list. I proposed a tier B+ for added flexibility, but no one else apart from Luke supported it...

So adding an S and B+ (or A-) tier seems like a good idea. 

39 minutes ago, LukeSavenije said:

so tldr: we can make a list for you that has 300 PSUs in it, all in their exact level, and we would have 300 tiers

That's taking the extreme view in an attempt to make the other side look crazy, we're not asking for 300 tiers, just a few more logical ones. Though if you went to spreadsheet form the 300 tier list would actually make sense and be readable. 

24 minutes ago, LukeSavenije said:

sorry, i don't have 5000 dollars spare for a load tester, a soldering iron, a oscilloscope, some PSUs, a designer so i can actually make money off of this and i don't think cm would let a 17 year old test a psu on their testing department

I'm pretty sure JonnyGuru has all or most of that and has been testing PSUs for... probably as long as you've been alive, but y'all did your best to run him off the forum.... 

EDIT: aight so he's been going at it since 2005, you would've been 3 when he started, going off the age you gave. 

[hello_there_123]

1 minute ago, Zando Bob said:

Seems like that'd make more sense, isn't there a risk rn of someone buying an RMi and thinking they're getting the same level of PSU as an AXi? 

Well... Yes...

2 minutes ago, Zando Bob said:

(can I get a quick explanation of the single rail vs multiple rail thing that actually makes sense instead of being a load of clumsily worded technical jargon designed to just confuse the person reading it instead of informing them? What I just described is all I've ever gotten from the self proclaimed PSU guys here)

https://www.overclock.net/forum/31-power-supplies/761202-single-rail-vs-multi-rail-explained.html

From 2010, still relevant

2 minutes ago, Zando Bob said:

So adding an S and B+ (or A-) tier seems like a good idea. 

I support the addition of those two, but getting everyone else onboard ain't easy...

 

3 minutes ago, Zando Bob said:

I'm pretty sure JonnyGuru has all or most of that and has been testing PSUs for... probably as long as you've been alive, but y'all did your best to run him off the forum.... 

Honestly, Wolf/whoever does the reviews now at JG.com, their reviews aren't that great as all they test are voltage regulation and ripple. Which even most low-end PSUs will do fine on. Transient response time, Protections, Noise, overall build quality etc. are what they usually fail at. 

[Zando_]

6 minutes ago, hello_there_123 said:

https://www.overclock.net/forum/31-power-supplies/761202-single-rail-vs-multi-rail-explained.html

From 2010, still relevant

thancc, I have consumed the informations. So basically the only difference is at what point it trips OCP, total load vs individual rail load from what I'm reading. 

[LukeSavenije]

5 minutes ago, hollyh88 said:

okay but those issues only have been created once you come close to the max output of the wattage right

correct, 120% to be exact.

[LukeSavenije]

32 minutes ago, Zando Bob said:

I'm pretty sure JonnyGuru has all or most of that and has been testing PSUs for... probably as long as you've been alive,

and hasn't been for years now...

 

I'll look into what i can do to change it here, maybe I'm even going with a score system

 

just keep in mind that can cost days, if not weeks or months to prepare and bring out

[Zando_]

1 minute ago, LukeSavenije said:

and hasn't been for years now...

Imma assume he still has knowledge on the subject, PSUs haven't radically and suddenly changed from what I've seen, a lot of the units were designed years ago. 
 

2 minutes ago, LukeSavenije said:

I'll look into what i can do to change it here, maybe I'm even going with a score system

 

just keep in mind that can cost days, if not weeks or months to prepare and bring out

I mean... you guys claim to be an authority on PSUs, I assume you know what goes where already? Should be a pretty quick edit to shuffle a few PSUs around in the list to make it more accurate while you deliberate on what kind of system you want to go with, then implement that. In the current state the list is in it's pretty sad to keep pointing newbies to it as an accurate resource, as I pointed out you'll easily have people thinking the RMi is on the same level as the AXi. Or the GQ is just as solid as a T2 (which from the Tom's Hardware - they do test protections and transient response and everything - was only beaten by the AXi at the time of review). Or any other number of PSUs on the same line as units much, much better than them. 

 

Right now the PSU list is full of misinformation unless you've already seen the old tier list and such and can mentally filter it. 

[LukeSavenije]

13 minutes ago, Zando Bob said:

T2 (which from the Tom's Hardware - they do test protections and transient response and everything - was only beaten by the AXi at the time of review)

but what aris doesn't mention there is that it's a non-configurable single rail (superflower, what a surprise). i can rule out uvp and I'll melt the cables with over 2000w of current. that's why i consider the flextronics axi better than the t2

 

I'll be working on a concept for the group tonight, which might see the light some day... give me time

[Zando_]

7 minutes ago, LukeSavenije said:

but what aris doesn't mention there is that it's a non-configurable single rail (superflower, what a surprise). i can rule out uvp and I'll melt the cables with over 2000w of current. that's why i consider the flextronics axi better than the t2

 

I'll be working on a concept for the group tonight, which might see the light some day... give me time

Yee I know that, I read the article @hello_there_123 linked me and see why multi rail is much safer on high wattage units. It defo loses to the AXi, buuuuut I don't think it's so far down that it should be in the same row with the GQ. Both the P2 and the T2 are farther right on the line than the GQ and such, reading from left to right as people reading english usually do, it seems like the farther left PSUs are the better choice. 

Basically if you're a noob and just reading the list without much knowledge of the PSU models and their features, it's easy to read the list as GQ > T2/P2. 

[LukeSavenije]

5 minutes ago, Zando Bob said:

Basically if you're a noob and just reading the list without much knowledge of the PSU models and their features, it's easy to read the list as GQ > T2/P2. 

I'll look what i can do

 

but we're of course talking about two completely different manufacturers here. superflower aims for performance, outsources a lot to rsy, while flextronics do (almost) everything in house, and because they're in origin a server psu manufacturer, focus more on stability

[Zando_]

1 minute ago, LukeSavenije said:

but we're of course talking about two completely different manufacturers here. superflower aims for performance, outsources a lot to rsy, while flextronics do (almost) everything in house, and because they're in origin a server psu manufacturer, focus more on stability

That.... has nothing to do with my questions and issues with the tier list? That doesn't even make sense because isn't stability a metric of performance for PSUs? Thanks for the tidbit I guess? 

[LukeSavenije]

8 minutes ago, Zando Bob said:

That doesn't even make sense because isn't stability a metric of performance for PSUs?

I'm talking about electrical performance, not stability

 

or in simpler terms: let it look good for reviews

[hollyh88]

1 hour ago, LukeSavenije said:

correct, 120% to be exact.

okay so as i stated with the parts i will be using i will never reach that wattage amount. so nothing to worry about and those psu's are then amazing? but why is everybody saying that they are bad then? when only thing you need to do is basically common sense.. not go over the total wattage that a psu can provide. I'm not an expert on psu's but even i know that  

[LukeSavenije]

7 minutes ago, hollyh88 said:

okay so as i stated with the parts i will be using i will never reach that wattage amount

well... with peak power of some GPUs it actually is possible and even doable. but the 2070 isn't one of them

 

8 minutes ago, hollyh88 said:

those psu's are then amazing

i find focus rather loud when the fan kicks in, and as many know I'm not a big fan of single rail units at that wattage

 

and even then, seasonic mostly designs, as has had enough with me that I'll just wait for the new ones, which they recently showed off

[hollyh88]

16 minutes ago, LukeSavenije said:

well... with peak power of some GPUs it actually is possible and even doable. but the 2070 isn't one of them

i mean 2080ti i suppose  
But i dont think a 2070 super/2080 would hit it even when overclocked.