[GamersNexus] How much wattage does your PC and your GPU actually need?

Page 2 - Seeking answers? Join the AnandTech community: where nearly half-a-million members share solutions and discuss the latest tech.


Diamond Member
Dec 7, 2003
I skimmed the article, but wanted to see a chart where they take the same system and just swap out the PSU to see how that affects power consumption at the wall. Like compare a no-name to bronze, silver, gold, platinum efficiencies to see how much you save.

As overall typical usage system power consumption goes down, I think it becomes less important to have a super efficient PSU, or at least it doesn't make sense spending huge amounts of money on it, because even if your PSU is inefficient, when you are talking about overall consumption of less than 100 watts most of the time, it would take years and years to pay off the difference in price between PSUs?

Yeah, for a typical desktop PC it's going to take a long time to pay off the difference if it's large. Heavy usage would obviously make the payoff faster, and might make it worthwhile. Even if you're running at 400W for 8 hours a day and 80W idle for the remainder (DC, not at the wall), the difference isn't huge.
Capstone 750W Gold 81.8%@79.4W, 90%@448W would be 5.12kWh/day
Corsair CX750M Bronze 78.7%@76.1W, 86.3%@376W would be 5.33kWh/day

Over the course of a year, that kind of high load would be 78kWh, which at $0.15/kWh is $11.70. Even at 2 hours at 400W and 22 hours at 80W per day, it's still 45kWh
Rosewill's updated the Capstone so it's not the exact same one in the review, but considering you can buy it for $55 after MIR, buying a $40 Bronze PSU like the EVGA 600B might not be worthwhile.

The reason why I'd spend more on a good Gold+ supply isn't so much the power savings, it's the other benefits like better cabling, lower fan noise, hybrid fan modes, etc.


Diamond Member
Mar 18, 2003
3/5 rating on new egg and with the experience I had for the past 3 years I would give it a 2. It is my second RMA. Been shutting down for a while even during light load. It is a whinning unit.

Didnt like it much except for its small form factor.

The EVGA superNova 1600 is freaking awsome and solid.

Customer reviews for PSUs are essentially useless, all you can really tell is if it provides power or not. It requires expensive equipment to properly test a PSU.


Platinum Member
Feb 24, 2005
There's also the issue of peak efficiency.

With this 1000W PSU you need about 500W load to reach peak efficiency:

With this 500W PSU you hit peak efficiency at around 300W (more typical for a mid-range/high-end PC with a single GPU):

Efficiency varies between different PSUs obviously, but the trend holds true. Higher rated PSUs need a heavier load before they can begin to reach peak efficiency, making lower rated PSUs more efficient for mid-range builds.

Consider that your computer probably spends more time at low-medium power than at 100% GPU and CPU load (browsing the web, media, light gaming) and an even lower rated PSU might be more efficient.

The system in my sig has been running with a Corsair 450VX which I kept from my previous Phenom II build. Absolutely no problems.
Last edited:


Golden Member
Jul 30, 2010
Customer reviews for PSUs are essentially useless, all you can really tell is if it provides power or not. It requires expensive equipment to properly test a PSU.

They are not useless, they give a pretty good idea. Sorry but I was able to determine if my PSU was dead without the expensive equipment you are referring too. Of course it is not a thourough review...it is a user review but those users can share their experience, whether it is positive or not.


Jan 8, 2011
With everything overclocked like crazy. Numbers are watts from the wall.

Firestrike extreme combined test: 800w
Crysis 3: 750-800w
BF4 on 64 player server (default clocked GPU's): 700w
GTAV: 750w

Hugo Drax

Diamond Member
Nov 20, 2011
Yet a 400w can't run an ssd, r9 nano, i7 3770k regular motherboard. When a game is loaded.

They should try this test with a 400w powersupply and see if it can really run.

The amp spikes which are too short to measure with the wattmeter is the problem. All it takes is a quick spike to cause the psu to go into protect mode. Even though when you have a big powersupply working and it looks like the PC only uses 280w