Search In
• More options...
Find results that contain...
Find results in...

# [Guide] Is a Certified PSU worth it?

## Recommended Posts

Posted · Original PosterOP

First of all, this guide will be covering only the aspect of efficiency and power requirements.  If you are looking for a guide on quality, please look elsewhere.

This guide aims to answer "should I upgrade my PSU to a more efficient 80+ gold (or platinum, etc.) model, and if so, which one?"

If you would like to know more about this standard and what each one means for the rated efficiency, see this post here.

The Calculations

The cost each year of operating the computer will be given by the following equation in every situation:

So together we have

Generally speaking, if the yearly cost of operating the power supply is greater than the cost of a more efficient power supply over the number of years you expect to use it for, you should get a better one.  But how can we know if it is actually cheaper?  Time for some examples!

Consider a light user

• System components draw of 150 W
• Time spent on PC: 4 hours per day, 5 days a week, 52 weeks per year = 4 * 5 * 52 = 1040 hours per year
• Lets assume power costs \$0.12 per kWh, and that this user is comparing an 80% efficient PSU with a 90% efficient PSU.

Yearly cost of running 80% efficient PSU: (150 W / 0.80) x (1040 h) x (0.12 \$/kWh) x (0.001 kWh/Wh) = \$23.40

(That last part is added as a conversion from kWh to Wh)

Yearly cost of running 90% efficient PSU: (150 W / 0.90) x (1040 h) x (0.12 \$/kWh) x (0.001 kWh/Wh) = \$20.80

That's a yearly savings of \$2.60 (\$23.40 - \$20.80) which is hardly anything at all!  This shows that if you are a very light user, it's not likely worth upgrading your PSU, since the cost of a new one could be well over \$100, and you'd have to use the computer for decades to make that back!

Consider a heavy user

• System components draw of 450 W
• Time spent on PC: 24 hours per day, 7 days a week, 52 weeks per year = 24 * 7 * 52 = 8736 hours per year
• Lets assume again that power costs \$0.12 per kWh, and that this user is comparing an 80% efficient PSU with a 90% efficient PSU.

Yearly cost of running 80% efficient PSU: (450 W / 0.80) x (8736 h) x (0.12 \$/kWh) x (0.001 kWh/Wh) = \$589.68

Yearly cost of running 90% efficient PSU: (450 W / 0.90) x (8736 h) x (0.12 \$/kWh) x (0.001 kWh/Wh) = \$524.16

That's a yearly savings of \$65.52 which is considerably more significant!  This shows that if you are a heavy user, it could be worth upgrading your PSU, especially if your system draws even more power than this, and/or your current PSU is even less efficient, and/or power is even more expensive where you live.

If math is not your strong suit, you can use the spreadsheet I've attached below.  It comes pre-filled with the numbers for the light user example, and you can adjust them to match your situation.

If you can read this you're using the wrong theme.  You can change it at the bottom.

##### Share on other sites
On 6/1/2016 at 6:05 PM, Ryan_Vickers said:

That's a yearly savings of \$65.52 which is considerably more significant

which is probably the money you'd spend on getting a more efficient PSU anyways (bit of an exaggeration but still)

LTT's self-appointed memelord

Always up for a game of chess

See @STRMfrmXMN @Energycore or @Starelementpoke for PSU needs

See @deXxterlab97 for blue fetishes (if that's your thing)

@Implosivetech's got you covered for all your graphic design needs

And lastly, never put a CX green series PSU in a high end build https://youtu.be/Oybz5Q-If9M?t=7

##### Share on other sites
Posted · Original PosterOP
1 hour ago, wrathoftheturkey said:

which is probably the money you'd spend on getting a more efficient PSU anyways (bit of an exaggeration but still)

Exactly my point

If more efficiency would save you that much, better to just buy a better one since it will pay for itself in a year and save you money after that.

If you can read this you're using the wrong theme.  You can change it at the bottom.

##### Share on other sites

I want to add a few things before a smarmy person does it. (does that make me smarmy as well? :3 )

PSU efficiency ratings start at 20% load. So anything below that is not guaranteed to be at least the rating on the sticker. Except for 80 PLUS Titanium which is guaranteed to be 90% at 10% load. As an aside, desktop PSUs are typically most efficient at 50% load: https://en.wikipedia.org/wiki/80_Plus#Efficiency_level_certifications

##### Share on other sites
Posted · Original PosterOP
38 minutes ago, M.Yurizaki said:

I want to add a few things before a smarmy person does it. (does that make me smarmy as well? :3 )

PSU efficiency ratings start at 20% load. So anything below that is not guaranteed to be at least the rating on the sticker. Except for 80 PLUS Titanium which is guaranteed to be 90% at 10% load. As an aside, desktop PSUs are typically most efficient at 50% load: https://en.wikipedia.org/wiki/80_Plus#Efficiency_level_certifications

Yes, good thing to keep in mind.  For the very light user who's system might never draw more than 200W, a 1500 W PSU could actually be less efficient than a smaller one, even if the smaller one isn't rated as well.

If you can read this you're using the wrong theme.  You can change it at the bottom.

##### Share on other sites

You should pin your own post.

Helios EVO (Main):

Intel Core™ i7-8700K (Won at LTX, signed by Dennis) | 16GB G.Skill Ripjaws V DDR4-3200 | GIGABYTE Z370 AORUS GAMING 5 | Sapphire AMD RX VEGA 64 | Define C TG -> NZXT H500 | Corsair CX550M | Cooler Master Hyper 212 BLACK RGB

Pacific Spirit XT (CPU signed by Dennis!) (Planned)

Spoiler

Intel Core™ i7-4720HQ (Taobao Special) | Asus MAXIMUS VI Formula | 16GB Kingston Black DDR3-1600 | 8GB G.Skill Ripjaws IV DDR3-1600 | ASUS DirectCU II HD 7970 3GB | Define C TG | Corsair CX550M | Intel Stock Cooler

Laptop (Legion Y7000)

Spoiler

i5 8300H | 16GB RAM | 256GB SSD | 1TB HDD | 1050 Ti | Power Brick

Intel is bringing DDR4 to the mainstream with the Intel® Core™ i5 6600K and i7 6700K processors. Learn more by clicking the link in the description below.