[EOL] PSU Tier List rev. 14.8

[LienusLateTips]

13 minutes ago, Cophee said:

Why are the Seasonic Focus Plus series not on the list? Is it cause they're being replaced by the GX/PX series? Did the focus plus units sit in A tier before being replaced?

They're on the list. Tier D+.

 

[CarbonVacuum]

On 12/29/2019 at 12:26 AM, Meldarion said:

Well I wasn't I was mainly hoping to hear someone compare the ax850 & prime ultra 850 to the t2 850 but everyone wanted to inform me that I'm making ridiculous selections. 

But since I have you here, I just completed a PCpartpicker list, my estimated wattage pre overclock and minus a handful of USB devies is 539, would you recommend a 550w PSU for that range?

 

edit: forgot to put in m.2 drives 559* estimated pre/oc

In my opinion, I wouldn't necessarily listen to the hardliner people who seem to be obsessed with informing thew world that they are selecting to high a wattage of power supply. Not saying they are 100% wrong. But they can't wait to tell people that 450w, or maybe 550w, is enough for their build. Sure, it often is. It also is not in many cases as well. When said build is something like yours, where you are powering many more devices than the bare bones PC user, more headroom can be desirable for fan noise, efficiency, longevity concerns... Many fans, lots of RGB, high end sound cards, many usb devices, many hardrives and ssd drives, a high wattage video card like a Sapphire Nitro+ 5700 XT, an overclocked Ryzen 9....

 

If your estimated wattage is 559w, do you really want to buy a 550w power supply? Yes it could work... I sure wouldn't want to try that. Why would anyone want to run their PSU at full load while gaming, or rendering etc.? Yes, you could get a 650W in that scenario. The fan would be louder than some people like. IT would not be running at peak efficiency, but I'll admit it's true that the efficiency at 90% load is still impressively high on modern, well built PSUs.

 

Get a 750w if you have the money. Hell, get an 850w. If your estimated wattage is well over 500, some people like their PSU to run at peak efficiency. That's usually around 50% load.

 

That's just my opinion. You could always just get a 550w. Save money and run that baby at full tilt for years while gaming or other high stress loads.

 

My recent build has a Seasonic GX-650. My estimated wattage is 558. I wish I bought a 750, but it's obviously fine as it is. But it's cutting way closer than I originally intended.

 

PCPartPicker Part List

Type	Item	Price
CPU	AMD Ryzen 5 2600X 3.6 GHz 6-Core Processor	$177.25 @ shopRBC
CPU Cooler	Cooler Master MasterAir G100M RGB 22.63 CFM CPU Cooler	$39.99 @ Amazon Canada
Motherboard	Asus ROG STRIX B450-F GAMING ATX AM4 Motherboard	$139.99 @ Memory Express
Memory	Team T-Force Delta RGB 16 GB (2 x 8 GB) DDR4-3000 Memory	$94.99 @ Newegg Canada
Memory	Team T-Force Delta RGB 16 GB (2 x 8 GB) DDR4-3000 Memory	$94.99 @ Newegg Canada
Storage	Corsair MP510 480 GB M.2-2280 NVME Solid State Drive	$115.99 @ PC-Canada
Storage	Western Digital Red Pro 10 TB 3.5" 7200RPM Internal Hard Drive	$346.99 @ Newegg Canada
Storage	Western Digital Red Pro 10 TB 3.5" 7200RPM Internal Hard Drive	$346.99 @ Newegg Canada
Video Card	Sapphire Radeon RX 5700 XT 8 GB NITRO+ Video Card	$579.99 @ Newegg Canada
Case	Cooler Master MasterBox NR600 (w/o ODD) ATX Mid Tower Case	$93.28 @ Vuugo
Power Supply	SeaSonic FOCUS Plus Gold 650 W 80+ Gold Certified Fully Modular ATX Power Supply	$129.50 @ Vuugo
Sound Card	EVGA NU 24-bit 192 kHz Sound Card	$282.68 @ Newegg Canada Marketplace
Case Fan	Noctua NF-S12A PWM 120 mm Fan	$27.96 @ Mike's Computer Shop
Case Fan	Noctua NF-S12A PWM 120 mm Fan	$27.96 @ Mike's Computer Shop
Case Fan	Noctua NF-S12A PWM 120 mm Fan	$27.96 @ Mike's Computer Shop
Case Fan	Noctua NF-S12A PWM 120 mm Fan	$27.96 @ Mike's Computer Shop
Case Fan	Noctua NF-S12A PWM 120 mm Fan	$27.96 @ Mike's Computer Shop
Case Fan	Noctua NF-F12 industrialPPC-3000 PWM 109.89 CFM 120 mm Fan	$39.99 @ Newegg Canada
Keyboard	Razer Cynosa Chroma Wired Gaming Keyboard	$54.99 @ Amazon Canada
Mouse	Logitech G300S Wired Optical Mouse	$38.99 @ Amazon Canada
Speakers	Edifier R1700BT 66 W 2.0 Channel Speakers	$179.99 @ Amazon Canada
	Prices include shipping, taxes, rebates, and discounts
	Total	$2896.39
	Generated by PCPartPicker 2019-12-30 13:08 EST-0500

 

 

Edited December 30, 2019 by CarbonVacuum
spelling

[LogicWeasel]

6 minutes ago, CarbonVacuum said:

My recent build has a Seasonic GX-650. My estimated wattage is 558. I wish I bought a 750, but it's obviously fine as it is. But it's cutting way closer than I originally intended.

You wrote a long post explaining your position on PSU use and how well reasoned it is to get a higher wattage power supply.  But then you lost me because your only "source" for how much power that system would likely use day-to-day is PCPartPicker ... lol.

 

PCPartPicker is not a precise tool to measure power-draw while gaming, it's just spitting the advertised TDP numbers at you (and in some cases numbers it makes up that often aren't fact-checked by its own website staff).  It thinks those HDD will be using 20W while the motherboard is pulling 70W (not for CPU) for example.  There are better ways to figure out how much power you need

[CarbonVacuum]

Yes, that's my source. I also used 7 PSU calculators.

 

I did not link those. But, now you know.

 

PC part picker just lists all the components nicely in a way I can link. I am not a professional, so I can't validate which one is supreme.

Edited December 30, 2019 by CarbonVacuum
added words

[LienusLateTips]

3 minutes ago, CarbonVacuum said:

Yes, that's my source. I also used 7 PSU calculators.

 

I did not link those. But, now you know.

 

PC part picker just lists all the components nicely in a way I can link. I am not a professional, so I can't validate which one is supreme.

PSU calculators are all completely bullshit. Your system won't be using more than about 450W. Maybe 500W during startup.

[Mezoxin]

14 minutes ago, CarbonVacuum said:

 

 

My recent build has a Seasonic GX-650. My estimated wattage is 558. I wish I bought a 750, but it's obviously fine as it is. But it's cutting way closer than I originally intended.

 

PCPartPicker Part List

PSU calculators and PC part picker takes into consideration full system load (except for  cpu running certain instruction sets )  and adds a safety factor ,   there is no real use scenario that would ever put your system in that kind of load , and if you reach that point of utilising every bit of resource your pc has , your system will be unresponsive and will require a hard reset ,  PC part picker tells me my estimated usage is 420w , but when gaming the maximum i have seen from my PSU monitoring chip is 330w  

[CarbonVacuum]

So in other words, the same rough amount that I found from the calculators. give or take 50 watts.

 

If they are bullshit, what do you guys use to arrive at that number in a speedy fashion? I assume just the manufactures listing for each item.

 

Also, if they went over by 50 - 75 watts, I am not too broken up about buying the GX-650. Even if my max draw is 500, 650 seems pretty reasonable. would not want less.

[LukeSavenije]

1 minute ago, CarbonVacuum said:

If they are bullshit, what do you guys use to arrive at that number in a speedy fashion?

wattages from cpu+gpu+50w=my number of choice

 

then i won't go borderline, like saying your system on a 450w, but a 550w wouldn't be a problem at all

[CarbonVacuum]

Peak efficiency and fan noise are ideal at 50% load. Some people care about that stuff, while also going with the valid presumption that a PSU operating at peak efficiency will have better odds for excellent or more longevity than a PSU operating at 95% load capacity. for years on end. There are no guarantees with electronics and longevity. There are only ways that you can reasonably increase your chances at a long life for your components.

 

I would rather run my PSU at 50% peak load when gaming, and I am already well past that point. That's just my opinion, and I am certainly not the only one. Not by a long shot. It just seems more rational, purely based on numbers. I got 12 years ( and counting ) out of a low end cooler master 500w PSU, using this rational as a guide. I see no reason to change that mind set. 

 

Even after reading every page in this, and other threads.

[LukeSavenije]

6 minutes ago, CarbonVacuum said:

Peak efficiency and fan noise are ideal at 50% load.

which is proven to be bs because modern llc resonant PSUs have a similar efficiency between 30 and 80%

 

fan noise depends heavily on fan curve

[CarbonVacuum]

2 minutes ago, LukeSavenije said:

which is proven to be bs because modern llc resonant PSUs have a similar efficiency between 30 and 80%

 

fan noise depends heavily on fan curve

Similar is not exact. I know the numbers. Also 80% load and 95% load are two sort of different loads that have a bearing on what I am saying in previous comments.

 

And yes, fan noise depends a lot on fan curve. I am aware.

[Mezoxin]

2 minutes ago, CarbonVacuum said:

Peak efficiency and fan noise are ideal at 50% load. Some people care about that stuff, while also going with the valid presumption that a PSU operating at peak efficiency will have better odds for excellent or more longevity than a PSU operating at 95% load capacity. for years on end. There are no guarantees with electronics and longevity. There are only ways that you can reasonably increase your chances at a long life for your components.

 

I would rather run my PSU at 50% peak load when gaming, and I am already well past that point. That's just my opinion, and I am certainly not the only one. Not by a long shot. It just seems more rational, purely based on numbers. I got 12 years ( and counting ) out of a low end cooler master 500w PSU, using this rational as a guide. I see no reason to change that mind set. 

 

Even after reading every page in this, and other threads.

PSU operating at peak efficiency does not affect the longevity of your components , actually if you overpower your pc with something like a 1000w PSU you are going to have horrible efficiency for most of your pc's usage time , as most systems up time is at Idle since you will be running it most of the time at 5-10% load which efficiency is at its worst 

 

[CarbonVacuum]

5 minutes ago, Mezoxin said:

PSU operating at peak efficiency does not affect the longevity of your components , actually if you overpower your pc with something like a 1000w PSU you are going to have horrible efficiency for most of your pc's usage time , as most systems up time is at Idle since you will be running it most of the time at 5-10% load which efficiency is at its worst 

 

I am not suggesting a 1000w PSU for a 300W system. I am suggesting peak efficiency, which I already stated.

 

In terms of longevity, I am unaware of the science that says running at varying efficiencies , do not effect the longevity of the PSU.

 

If that is the case, than I am happy to learn those materials. There is also the matter of just wanting to run at peak efficiencies, for the satifaction, and the tiny power bill adjustment. but I assume longevity as well, until I learn otherwise. It's worked for me so far for 35 years of PC use.

Edited December 30, 2019 by CarbonVacuum
words added

[Mezoxin]

4 hours ago, CarbonVacuum said:

I am not suggesting a 1000w PSU for a 300W system. I am suggesting peak efficiency, which I already stated.

my system with an i7 9700k and RTX 2070 idles at 60-70w during this power draw i will be  listening to music watching movies browsing the internet using  microsoft office  your system will be about the same as well , if you buy a 750w PSU for that system it means your system will be spending most of its time at a maximum of  8% load and efficiency  would be lower than what you would experience if you have a 550w psu 

But again efficiency doesn't really matter as we are speaking difference of a couple of percentages , and your actual gaming load with your system would be 60-70% of what PSU calculator suggested you , because as i mentioned earlier there is no real case scenario that will consume all your system resources at the same time 

[CarbonVacuum]

I don't have the same use case as you.  I have a 650w PSU. In light of my use cases, and my components, it was the correct purchase. The Seasonic GX-650.

 

It's good that I won't use all my resources at the same time. That way, I am getting even closer to the previously mentioned peak efficiency under load.

 

Also, there is overclocking. That really ramps up the power needs

Edited December 30, 2019 by CarbonVacuum
oc

[Mezoxin]

19 minutes ago, CarbonVacuum said:

Also, there is overclocking. That really ramps up the power needs

Depends on the system components , when a person OC's as a rule of thumb he may add 100W , but some configuration like what you own for example the 2600x there is little headroom left untapped for overclocking  and while they only consume an average of 30w higher when oc'ed they dont really offer a noteworthy improvement in performance , so normally a person who buys a 2600x doesn't really care about overclocking and even if they do their OC wont cause a significant  increase with cpu power consumption as with something like intel 9700k or 9900k 

[LukeSavenije]

36 minutes ago, CarbonVacuum said:

In terms of longevity, I am unaware of the science that says running at varying efficiencies , do not effect the longevity of the PSU.

so what you're saying is that a lower efficiency psu would die earlier?

 

because that one i have never heard before

 

and yes, there can be given a relation to the load it runs at, but not directly. this would be mostly with the heat it runs at, which can be changed by many other factors

 

and again, 50% is generally not even the actual peak

[CarbonVacuum]

I got my 2600x for $159 cdn. $5 more than the 2600. I am interested in OC. Little head room there may be. I was originally going to buy the 2600 but for $5 more, I said why not. Plus I use my PC at high power draw a lot, and its on all the time.

[CarbonVacuum]

2 minutes ago, LukeSavenije said:

so what you're saying is that a lower efficiency psu would die earlier?

 

because that one i have never heard before

 

and yes, there can be given a relation to the load it runs at, but not directly. this would be mostly with the heat it runs at, which can be changed by many other factors

 

and again, 50% is generally not even the actual peak

What I am saying is, where is the science that say load and efficiency doesn't matter for longevity?

 

And I've read several times that 50% load is peak efficiency. If you're saying that's wrong, then I would like to see the science on that as well.

 

I am happy to learn.

[LukeSavenije]

6 minutes ago, CarbonVacuum said:

And I've read several times that 50% load is peak efficiency. If you're saying that's wrong, then I would like to see the science on that as well.

to show some examples of this (randomly chosen)

 

[CarbonVacuum]

I have seen these and other charts. Much less importantly for me - It doesn't say that the efficiency is the same for all loads or loads ouutside of and somewhat close to 50%, much more importantly - nor does it indicate how loads and efficiency effect longevity. If you have those research papers, I will gladly read them.

 

 

Edited December 30, 2019 by CarbonVacuum
added words

[Mezoxin]

  

11 minutes ago, CarbonVacuum said:

What I am saying is, where is the science that say load and efficiency doesn't matter for longevity?

 

 

efficiency simply gives an indication of how much electricity is wasted as heat 

Can heat affect the lifespan of components ? Definitely yes and it was a big deal a couple of decades ago 

But .. Modern PSU's are very efficient  that at the worst point of their  efficiency curve they waste a very small amount of energy as heat , which makes PSU efficiency nowadays  to be negligible in affecting the longevity of any modern properly designed PSUs while operating under its operating limits 

8 minutes ago, CarbonVacuum said:

I have seen these and other charts. Much less importantly for me -

the measurements you see in these charts were made by  an electrical engineer with degrees in Computer Science, Telecommunications, and a Ph.D in the field of wireless communications and has worked on the development and testing of countless power supply units in both industrial and media related positions. He is also the lead PSU reviewer for both Tom’s Hardware and TechPowerUp

[CarbonVacuum]

2 minutes ago, Mezoxin said:

  

efficiency simply gives an indication of how much electricity is wasted as heat 

Can heat affect the lifespan of components ? Definitely yes and it was a big deal a couple of decades ago 

But .. Modern PSU's are very efficient  that at the worst point of their  efficiency curve they waste a very small amount of energy as heat , which makes PSU efficiency nowadays  to be negligible in affecting the longevity of any modern properly designed PSUs while operating under its operating limits 

the measurements you see in these charts were made by  an electrical engineer with degrees in Computer Science, Telecommunications, and a Ph.D in the field of wireless communications and has worked on the development and testing of countless power supply units in both industrial and media related positions. He is also the lead PSU reviewer for both Tom’s Hardware and TechPowerUp

And? Where are the research papers that indicate how loads and efficiency affect longevity? I will happily read them.

[LukeSavenije]

16 minutes ago, CarbonVacuum said:

If you have those research papers, I will gladly read them.

it was to show 50% is almost never the highest point. people just think so because 80+ measures only at (10% for titaniun), 25%, 50%, 75% and 100%

 

i don't have papers on it, but part of it makes sense. the more stress you put on components, the hotter it gets. but this is for example also affected by outside temperature

 

that's why I'd say it's hard to say if it'll actually affect

[CarbonVacuum]

those charts seem to indicate peak efficiency is around 50% or 55%, and that that efficiency goes down with higher loads. Which is what I was already indicating. 

 

Plus, where are the long term research papers regarding loads and efficiency affecting longevity? Until those say that loads and efficiency do not alter longevity, I will continue to not want to tax my PSU to a huge degree. That's just my outlook. I like to go with what I see as being the most reasonable, until better science forces my view to change.

 

Not sure why wanting to see the science on this seems to be contentious. I am happy to learn, if these claims are based in science. Also, there are 2 different matters being discussed here.