which bang for your buck 80+ semi modular PSU would suit my PC?

[DeadnightWarrior]

This is, more or less, my PC:
https://it.pcpartpicker.com/list/gDGrCz
You can also find the specs on my LTT forum profile.

Note that the CPU cooler I'm using is the AMD Spire, not the one you see in the list.
 

It is currently powered by a Thermaltake Hamburg 530W 80+ unit that's been serving me well for the past 6+ years.

I'd like to put this unit in a retro gaming PC I'm building and get a semi modular PSU for my main rig.

Please note:
- I never overclock anything: CPU, RAM and GPU are all working at stock speed and I have no intention to change that.

- I couldn't care less about RGB, AIOs, extensions, a million fans and such. I have an air cooler, two intake fans and one exhaust fan. Period.

- I don't want to overspend, I'm looking for something that's right for the job and MAYBE allow for a future CPU / MB upgrade (always in Ryzen 5 / i5 and mid range GPUs territory, no more than that).

- I honestly think fully modular PSUs are useless (you HAVE to plug those 24 pin, 4+4 pin and PCI-E cables, right?!), this is why I'm only focusing on semi modular units: all I need is the mainboard, CPU, GPU links and two SATA connectors.

 

That being said, PcPartPicker estimates my total wattage to be 364W; this made me focus on 450W to max. 550W units.

 

I've narrowed my choices down to these models:

- BeQuiet System Power 9 CM 500W (80+ Bronze)
- BeQuiet Pure Power 11 CM 500W (80+ Gold)

- Corsair CX450M (80+ Bronze)

- Cooler Master MasterWatt 550W (80+ Bronze)
- Sharkoon WPM Gold Zero 550W (80+ Gold)
- Seasonic CORE GM-500 (80+ Gold)

 

Out of these, the Pure Power 11 is the most expensive at around 80€, while the Sharkoon WPM is the cheapest one at around 66€.

Which one would you buy, which would you avoid and why?

 

Thank you!

 

[jaslion]

Best to worst:

Pure power 11

Seasonic core

Sharkoon gold

Cx450m

System power 9

Masterwatt

 

I would personally pick between the core or pure power as well everything below that is c tier stuff.

[Imglidinhere]

50 minutes ago, DeadnightWarrior said:

Snip

 

Worth noting that the price difference between a high quality 500w and 700w unit is basically only around $30~

 

Getting a higher wattage unit isn't a bad idea when it's that close for pricing.

[DeadnightWarrior]

1 minute ago, Imglidinhere said:

 

Worth noting that the price difference between a high quality 500w and 700w unit is basically only around $30~

 

Getting a higher wattage unit isn't a bad idea when it's that close for pricing.

well, let's say I'm fairly sure I would never really use 700W of max power (unless I bought a very high end GPU which is never going to happen )
My starting point was around 50€, go figure. By adding the semi modular option I already raised the bar to around 70€, which is dangerously close to what I'd ever be willing to spend...

[IIIIIIIIII]

Of the products listed, only the Seasonic Core or Sharkoon WPM is worth it.

 

If you can add 10 euros, get a much better product, Whisper M.

 

Formula Gold is recommended if you don't mind if it's a non-modular product, and MWE Bronze V2 (but only for 100V to 240V versions) is fine if you're looking for something a little cheaper.

[Imglidinhere]

7 hours ago, DeadnightWarrior said:

well, let's say I'm fairly sure I would never really use 700W of max power (unless I bought a very high end GPU which is never going to happen )
My starting point was around 50€, go figure. By adding the semi modular option I already raised the bar to around 70€, which is dangerously close to what I'd ever be willing to spend...

It's nothing to do with "I would never use that much power" and more a matter of... why shortchange yourself? If you're intentionally limiting your PSU because you're trying to save pennies, that's silly. If you can't afford to purchase something more expensive, that's fine but that needs to be stated. I'm under the assumption that you had no budget since nothing was listed.

 

Also, PSUs are generally at maximum efficiency when they're at approximately 50% load. So if you're going to be pushing 365w through a 500w unit, you're outside the top efficiency curve. JohnnyGuru tests this sort of thing extensively. That and lower stress loads on a PSU tend to help it last longer.  Hope this helps.

[gloop]

3 hours ago, Imglidinhere said:

Also, PSUs are generally at maximum efficiency when they're at approximately 50% load.

Was that not disproven a while ago or am I misremembering?

[Imglidinhere]

2 hours ago, gloop said:

Was that not disproven a while ago or am I misremembering?

Nope, it's 100% true. They reach optimal efficiency at around 50% load unless specifically designed otherwise. Plenty of reviews showcase this and it's not a particularly hard measurement to find. The tools needed to test and note the actual efficiency are relatively cheap too.

[seon123]

12 minutes ago, Imglidinhere said:

Nope, it's 100% true. They reach optimal efficiency at around 50% load unless specifically designed otherwise. Plenty of reviews showcase this and it's not a particularly hard measurement to find. The tools needed to test and note the actual efficiency are relatively cheap too.

Please at least have the decency to stop talking BS when someone calls you out for it...

 

 

First of all, the efficiency resembles a horizontal line, more than a curve. You would know this, if you had actually bothered to read any review. If you want a more efficient PSU, then get a more efficient PSU, not a higher wattage one. Looking at the seven most recent reviews by Aris on THW, we can see that their efficiency stays within a few percentage points, all the way from 20% load to 100%. We can also see that they are most efficient at 30%, 30%, 40%, 30%, 40%, 40%, and 40% efficiency. They range from a bottom of the barrel garbage, group regulated EVGA N1, to the fairly high end FSP Hydro PTM. Not that any of this matters, as efficiency is one of the least important factors when buying a PSU for a normal system.

 

In order to actually test the efficiency of a PSU, you would need a load tester, which by itself is a very expensive piece of equipment. As if we needed even more confirmation of where you're getting your information from.

[boghubodaghi]

6 hours ago, seon123 said:

First of all, the efficiency resembles a horizontal line, more than a curve. You would know this, if you had actually bothered to read any review.

Agree on most of the points that you make, but I disagree with this one. Efficiency on PSUs more resembles a curve than a horizontal line.

 

And if you're looking for an efficient PSU it depends on what you are looking for, if you're nitpicky then you can compare the data from Aris' reviews and gauge which one is more efficient and at what loads.

[Imglidinhere]

8 hours ago, seon123 said:

Please at least have the decency to stop talking BS when someone calls you out for it...

 

 

First of all, the efficiency resembles a horizontal line, more than a curve. You would know this, if you had actually bothered to read any review. If you want a more efficient PSU, then get a more efficient PSU, not a higher wattage one. Looking at the seven most recent reviews by Aris on THW, we can see that their efficiency stays within a few percentage points, all the way from 20% load to 100%. We can also see that they are most efficient at 30%, 30%, 40%, 30%, 40%, 40%, and 40% efficiency. They range from a bottom of the barrel garbage, group regulated EVGA N1, to the fairly high end FSP Hydro PTM. Not that any of this matters, as efficiency is one of the least important factors when buying a PSU for a normal system.

 

In order to actually test the efficiency of a PSU, you would need a load tester, which by itself is a very expensive piece of equipment. As if we needed even more confirmation of where you're getting your information from.

https://www.tomshardware.com/reviews/psu-buying-guide,2916-3.html#:~:text=Looking at the curve%2C we,50% of its nominal capacity.

 

https://www.anandtech.com/show/2624/debunking-power-supply-myths/3

 

https://adoredtv.com/how-to-choose-the-right-power-supply/

 

If you want to argue with Tomshardware, JonnyGuru, Anandtech, AdoredTV and a whole host of other top tier reviewers, please continue to tell me idiotic things like "You need to buy a higher rated PSU if you want higher efficiency!" No shit.

 

The OP is asking about an 80+ model to get, so it's already at minimum 80+ White, which is the absolute lowest 80+ rating. That efficiency is at it's highest point at 50% load. So for an 80 Plus Gold rated PSU at 700w, you will hit that 87% efficiency sweet-spot at around the power draw the OP says his system will pull. By all means, continue to tell me that the information I'm speaking is wrong.

[seon123]

4 hours ago, Imglidinhere said:

https://www.tomshardware.com/reviews/psu-buying-guide,2916-3.html#:~:text=Looking at the curve%2C we,50% of its nominal capacity.

 

https://www.anandtech.com/show/2624/debunking-power-supply-myths/3

 

https://adoredtv.com/how-to-choose-the-right-power-supply/

 

If you want to argue with Tomshardware, JonnyGuru, Anandtech, AdoredTV and a whole host of other top tier reviewers, please continue to tell me idiotic things like "You need to buy a higher rated PSU if you want higher efficiency!" No shit.

All you did there, was show that you don't read reviews, and are not able to differentiate between people spewing BS, and people that are knowledgeable

You also did nothing to address the numbers that I showed for the most recent reviews on THW.

4 hours ago, Imglidinhere said:

The OP is asking about an 80+ model to get, so it's already at minimum 80+ White, which is the absolute lowest 80+ rating.

If you had actually looked at what OP wrote, you would see that they were looking at 80+ rated PSUs. As in, they have an 80+ rating. Not that they were looking for 80+ White PSUs. This should be pretty obvious, if you had actually bothered to read anything.

 

 

Not to mention that you did nothing to address the last part, which called you out on this part:

13 hours ago, Imglidinhere said:

The tools needed to test and note the actual efficiency are relatively cheap too.

 

13 hours ago, seon123 said:

In order to actually test the efficiency of a PSU, you would need a load tester, which by itself is a very expensive piece of equipment. As if we needed even more confirmation of where you're getting your information from.

 

6 hours ago, boghubodaghi said:

Agree on most of the points that you make, but I disagree with this one. Efficiency on PSUs more resembles a curve than a horizontal line.

 

Do note that the vertical axis on the graphs don't start at 0% efficiency, and don't go up to 100%. They obviously do this to make the very small differences in efficiency visible. Just putting the numbers from the review into a proper graph shows.... that it's pretty much a horizontal line past 100W-ish. There are of course many different ways of showing the exact same data, but I feel like this is more representative, since we want to know how much the efficiency actually varies, rather than how it varies.

Spoiler

 

Edited March 21, 2021 by seon123
Spellling mitsake

[boghubodaghi]

42 minutes ago, seon123 said:

Do not[e] that the vertical axis on the graphs don't start at 0% efficiency, and don't go up to 100%. They obviously do this to make the very small differences in efficiency visible. Just putting the numbers from the review into a proper graph shows.... that it's pretty much a horizontal line past 100W-ish. There are of course many different ways of showing the exact same data, but I feel like this is more representative, since we want to know how much the efficiency actually varies, rather than how it varies.

  Hide contents

 

It's still a curve, even though the variation is not large. But contextually, I don't deny that you are right.

 

@Imglidinhere

The chance of having a more efficient PSU at a given load is pretty slim, as shown by the graphs showed. Probably only happens when you're comparing a Non-Titanium 1500+W PSU to a 450W PSU. In OP's case getting a better rated version will always grant better efficiency.

[DeadnightWarrior]

It's very interesting to follow this thread, I'm no PSU expert so it all helps!

My 2 cents about my particular case:

- I'm running my PC with a 6 years old 530W 80+ white PSU: a theoretical max load 365W / 220v would be equal to 365 / 0.8 = 456W pulled from the wall. That's 91W wasted.
- Keeping the exact same system but using a 500W 80+ gold PSU, I'd have 365 / 0.9 = 405W from the wall. That's 40W wasted.

 

So I guess saving up to 51W simply by swapping the power supply isn't that bad, is it?

In the end I gave in to the Seasonic Core GM, as nearly everyone I asked named it the best among my choices.

[--SID--]

Pure Power 11 has a better fan, is more silent, has better cables and better protections because every single pci-e cable (it has two with one 6+2 pin connector each) is on a different 12v rail with OCP on each rail. I know systems with RTX 3070 that runs flawless on a PP 11 500w.

[Imglidinhere]

Hilarious, people can't understand context and only hear what they want.

 

Breakdown tiiiiime:

 

1) OP was looking at an 80 Plus PSU.

2) I assumed every PSU he's looking at will be of an 80 Plus certification.

3) Given the overwhelming number of 80 Plus Gold PSUs he has listed, it's hard to find a PSU that DOESN'T meet this criteria and/or that he'd ditch the 80 Plus moniker and buy some no-name brand.

4) Using this line of thought, I recommended him to buy a higher wattage unit specifically because the price difference is not great enough to warrant spending less for a lower output offering unless his budget is that tight, under the assumption that he'd be buying an 80 Plus certified unit.

 

TA-DAAA!

 

Nothing in this chain of posts has me saying to ignore the 80 Plus cert and buy a higher wattage, non-80 Plus PSU. Not one. Stop putting words in my mouth. There's no hidden, subliminal messaging going on here. xD