Very low power consumption 150W max on gaming pc?!

[Mariachi]

According to reviews and info i got online and at power consumption calculator for my system i got 820Watt PSU, however when measured with P3 Kill A Watt Electricity Load Meter and Monitor it shows max power used by my computer to be ~130-180 Watt max under any load from game or benchmark i have tried. Meter is correct i have tested on number of appliances in my home and it shows precise power consumption.

Another thing between idle load to full load on CPU and VIDEO power consumption does not change a bit.

Specs:

• Intel Core i5-3570K (TDP 77 Watt)
• 2x Nvidia 660Ti graphic card in SLi mode (each TDP 150 Watt, 450 Watt Minimum PSU required by NVIDIA)
• 32GB of A-Data RAM
• Creative Sound Blaster Z
• 32GB (4 x 8GB) HyperX 240-Pin DDR3 1600
• HDD 1 SSI 256GB
• HDD 2 7200RPM 1TB
• 6 FANS in chassis
• Nexus 820Watt Silent Power Supply

What's the deal here are we being deceived by evil marketers that we need so much power when in reality i could be good with cheap 300W PSU.

 

[Dasa2]

some online calculators are crap
going by this your recommended psu is 550w without overclocking http://outervision.com/power-supply-calculator

going by this oc models of your gpu use ~130w each
http://www.techpowerup.com/reviews/Zotac/GeForce_GTX_660_Ti_Amp_Edition/26.html

some cheap power meters get thrown of by some devices and wont report correctly sounds like your psu is one of them
does it give a pfc reading?

 

[lehtv]

some online calculators are crap
going by this your recommended psu is 550w without overclocking http://outervision.com/power-supply-calculator

That's an example of a crap PSU calculator. With the OP's setup it gives you about 500W max load wattage - which is in the right neighborhood - but then it recommends a 550W PSU. No one in their right mind would buy a 550W PSU for a 500W load system, that'd just place an unnecessarily high load on the unit, causing it to heat up, make a lot of noise and possibly die a premature death. A 550W unit typically doesn't even have the 4 x 6-pin PCIe connectors required for 660 Ti SLI. 650-750W would be more appropriate.

What's the deal here are we being deceived by evil marketers that we need so much power when in reality i could be good with cheap 300W PSU.

You can be sure your PC is using power normally, and your Kill-A-Watt is showing incorrect readings.

 

[UsandThem]

What's the deal here are we being deceived by evil marketers that we need so much power when in reality i could be good with cheap 300W PSU.

I think your system should use a little more wattage than you're reading. I have an i7 and a GTX 970, and I've seen my go into the mid-to-upper 200's when gaming. I only went with a 650w unit because at the time, it was the same price and I figured I'd leave a little play room to see if the new GPUs coming this year were going to continue increasing in power requirements (especially on the AMD side).

Outside of that, I think a lot of comes down to marketing and computer people telling people they need a huge PSU (for future-proofing or it's only $20 more to get a bigger unit).

Not many people truly need: 32 GB of 'gaming' ram, GTX 980ti cards (1-3 SLI), PCIe hard drives, $400 motherboards, $200 sound cards, 1000w power supplies, 5930k CPUs, or half of the stuff we buy.

A companies job is to make stuff to sell to us, and if for some reason we aren't buying, to find some way to make us think we need it.

If I owned Intel, I'd rather more people bought more 5930k CPUs instead of Pentium G4400 CPUs. If I owned Seasonic, I'd rather people bought 1200w Platinum units instead of 500w Bronze units.

 

[Dasa2]

That's an example of a crap PSU calculator. With the OP's setup it gives you about 500W max load wattage - which is in the right neighborhood - but then it recommends a 550W PSU. No one in their right mind would buy a 550W PSU for a 500W load system, that'd just place an unnecessarily high load on the unit, causing it to heat up, make a lot of noise and possibly die a premature death. A 550W unit typically doesn't even have the 4 x 6-pin PCIe connectors required for 660 Ti SLI. 650-750W would be more appropriate.
.

the 550w recommendation is fine as a minimum provided the psu is decent quality and has the required connectors under typical use the system will use ~300-400w
650w would leave room for overclocking and some upgrades
750-850w is enough for 980ti sli edit: with 3570k<

my 2600k@4.3 r9 290@1100 system has similar power use to that system
although it can go much higher if i crank up the v for overclocking pushing to 5.1ghz 1.52v 1225 +200mv it can hit around 600w

 

[Hail The Brain Slug]

Part of the problem is there really aren't that many ways to fully load all your power-hungry components without running multiple stress tests at once (i.e. prime95 + a video card stress test)

If you haven't overclocked your stuff, you are also dealing with the TDP limits of your cards and your CPU.

For example: I have a 5960x @ 140w TDP and a 980ti @ 250 watt TDP. Under full load, I would expect them to use about that much power, and I could proably get away with a 600-650 watt PSU to power them.

I have my 5960x overclocked to the point it may draw 300-400 watts depending on the type of load, and I've flashed my 980ti to increase the TDP limit to 365 watts. If I run AVX prime 95 and uningine valley at the same time, I can observe upwards of 800 watts power usage.

800+ watts with all my overclocks, but if I load up Diablo 3 at 4k with vsync on, my system only draws 250-350 watts. Does that mean I could get away with a 400 watt PSU for gaming? No, because as soon as a part is stressed by something I do, it could easily jump way beyond the PSU specification.

 

[tortillasoup]

According to reviews and info i got online and at power consumption calculator for my system i got 820Watt PSU, however when measured with P3 Kill A Watt Electricity Load Meter and Monitor it shows max power used by my computer to be ~130-180 Watt max under any load from game or benchmark i have tried. Meter is correct i have tested on number of appliances in my home and it shows precise power consumption.

Another thing between idle load to full load on CPU and VIDEO power consumption does not change a bit.

Specs:

• Intel Core i5-3570K (TDP 77 Watt)
• 2x Nvidia 660Ti graphic card in SLi mode (each TDP 150 Watt, 450 Watt Minimum PSU required by NVIDIA)
• 32GB of A-Data RAM
• Creative Sound Blaster Z
• 32GB (4 x 8GB) HyperX 240-Pin DDR3 1600
• HDD 1 SSI 256GB
• HDD 2 7200RPM 1TB
• 6 FANS in chassis
• Nexus 820Watt Silent Power Supply

What's the deal here are we being deceived by evil marketers that we need so much power when in reality i could be good with cheap 300W PSU.

Test your power meter by putting it on a lamp socket and putting a known load like a 60w incandescent bulb. I've seen those P3 watt meters having inaccurate readings and had to replace one. I have about 4 of these meters.

 

[Dasa2]

Test your power meter by putting it on a lamp socket and putting a known load like a 60w incandescent bulb. I've seen those P3 watt meters having inaccurate readings and had to replace one. I have about 4 of these meters.

a 60w incandescent bulb will have a power factor close to 1 which makes it very easy for a cheap power meter to get a accurate reading a cheap psu or electric motor on the other hand can be much more difficult

so a meter that is accurate with one thing can still be way out on another

 

[tortillasoup]

a 60w incandescent bulb will have a power factor close to 1 which makes it very easy for a cheap power meter to get a accurate reading a cheap psu or electric motor on the other hand can be much more difficult

so a meter that is accurate with one thing can still be way out on another

This is true, which in that case, he can do a test both with an incandescent and a CFL/LED.

 

[darkfalz]

Many cheap power draw measurers don't properly measure switching power supplies as the duty cycle confuses them. For example a cheap one I had showed it using 60 watts in standby mode, when the actual usage was 4 watts.

That one was $20 - I'd say it's useless to measure the power draw of a PC. The cheap one I had was similar. I got another one for about $80 and it was accurate.

https://images-na.ssl-images-amazon.com/images/I/41aHr7NS9pL.jpg

This is a common one that is accurate, however the design may have since been taken and used for cheaper ones too.

 

[Yuriman]

My Kill-a-watt reads accurately - I've tested several several loads with varying power factors, and the numbers add up.

My entire i5 3570K system, at stock clocks, only draws 81.8w at the wall under a Prime95 load. Prime95 + Furmark is only 185w. If I OC the CPU to 4.5ghz, total power consumption goes up to 259w. I expect that, at stock, a 300w power supply would be plenty and then some, while overclocked I would want 350-400w to give plenty of headroom.

Power supply recommendations are usually a lot higher than is really necessary, I find.

 

[WackyWRZ]

Another thing between idle load to full load on CPU and VIDEO power consumption does not change a bit.

This is the key - the meter is not reading something correctly here. I see my system swing by at least 150 watts between load and idle when using my Kill-a-Watt.

 

[piasabird]

Are you counting the monitor, speakers, etc?

 

[piasabird]

You might try turning off the computer and turn it back on and see how much power it draws when it boots up.

Computer parts often use less power than their official rating.

 

[poohbear]

For maximum efficiency you should be running @ 50% of your maximum wattage. Just because u use 300wts doesn't mean a 350wt PSU is best for u. @ 300wt system usage, a 600wt PSU would provide you with the best power efficiency.

 

[poohbear]

Are you counting the monitor, speakers, etc?

Don't most monitors & speakers use their own PSU? They're not connected to the PC'S PSU so why would he factor that in?

 

[Yuriman]

For maximum efficiency you should be running @ 50% of your maximum wattage. Just because u use 300wts doesn't mean a 350wt PSU is best for u. @ 300wt system usage, a 600wt PSU would provide you with the best power efficiency.

50% is peak efficiency, yes, but if you only load @ 300w on a 600w power supply, partial loads and idle will be far less efficient because they're much farther from peak efficiency.

 

[piasabird]

Most power supplies in their technical schematics have what is called a Power curve. At lower levels of the curve and higher levels of the curve the power may be somewhat less stable. I would use the assumption that during the playing of an intense video game you will get a somewhat higher power usage. Peak power usage should be near the sweat spot of the curve. This keeps the power supply from working too hard or wearing out too soon or blowing out completely or also it may run even hotter than it needs and create heat instead if dissipating it through the cooling fins on the power supply.

A power supply is very complicated. It has several lines or rails of power being produced in several different voltages. Each of these lines can be stressed out individually. How they get the wattage rating is by adding up the wattage/amperage on each line. So you should take the total wattage rating on a power supply with a grain of salt. If one power line fails, they all fail. Not all power supplies actually can reach their peak power rating on all the power lines/rails. So it can be a safety factor to increase the wattage rating of the power supply a bit. Tom's Hardware has done some testing in this area and the results were not always good.

 

[Dasa2]

For maximum efficiency you should be running @ 50% of your maximum wattage. Just because u use 300wts doesn't mean a 350wt PSU is best for u. @ 300wt system usage, a 600wt PSU would provide you with the best power efficiency.

with most decent psu the efficiency drop as you go higher is minimal
here its just 1% up to 80% load it drops quicker under 30% load
for max efficiency i like to aim at ~80% for the systems max usage and ~50% for the real\typical usage if possible

noise is another matter if you want a quiet system psu fans often go up in noise a lot over ~40% load

 

[mnewsham]

I recently got a UPS unit that provides power draw of what I have plugged into it. My old computer (A10-7850k 2x r7-260x, 16GB 2133mhz DDR3 with a 550w 80+ bronze PSU) would idle around 130-150w and hit ~400w under load. My new computer (i7-5820k, GTX 960, 32GB 3200mhz DDR4 with a 600w Silverstone 80+ titanium PSU) idles around 80-100w and hits around 325w under load.

Part of the idle power draw is improved because of the 90% efficiency at 10% load the titanium 80+ spec requires. But I also suspect the newer gen parts are simple more power efficient than what I had previously.

Either way I'm very happy with my new computer, new UPS, and my new lowered idle power draw.