NEW STUDY: Taming the Energy Use of Gaming Computers

[nims0c]

If people would only actually buy adequate PSUs, maybe that would help.

 

The amount of dudes I see with 750w PSUs, but only 300W-ish configs is insane....

 

Aren't PSUs most efficient around 50% load?

[linuxfan66]

Gpus use more proportionally than they once did in upper end now. Last time I checked 2003 didn't have 300w consumer gpus..

[MEC-777]

If people would only actually buy adequate PSUs, maybe that would help.

 

The amount of dudes I see with 750w PSUs, but only 300W-ish configs is insane....

 

This is mostly irrelevant. A PSU will only output as much power as is required by the system. A 750w PSU is not pulling 750w from the wall at all times. It only pulls what the components need. 

 

Also, the majority of modern PSU's (anything 80+ bronze or better) is generally operating at near-peak efficiency across a large portion of it's load range (25-90% load).   

[Trik'Stari]

I feel like I'm the only one who could not care less about power consumption.

 

 

f**k this planet, I wanna live on mars lol

 

Plot twist, cows actually make a metric shit ton (no pun intended) of greenhouse gasses. Stop eating so many burgers to save the world

 

If we don't eat them, they'll overpopulate and produce more greenhouse gasses. Unless you're suggesting we just slaughter all of them needlessly to protect ourselves.

[MEC-777]

I feel like I'm the only one who could not care less about power consumption.

 

 

Don't worry, you're not the only one.

 

I think people get confused about this. "Power consumption" should not be a concern where performance is of utmost importance (when it comes to gaming), but "efficiency" should be a concern. More performance from the same or lower TDP GPU's is a very good thing, since it can lead to more overclocking headroom, less heat output, and the potential for even more powerful "bigger" GPUs. 

 

Unless someone has a legitimate concern with regards to power consumption (extreme high cost of local energy or small room that heats up easily, etc.), it should not be a deciding factor.

[Imakuni]

Aren't PSUs most efficient around 50% load?

Yup. As a generic rule, it works best at 40~60% loads. And to be even more precise, it depends on the model, but the peak efficiency is reached a bit over 50%, and a bit under 55% loads.

 

This is mostly irrelevant. A PSU will only output as much power as is required by the system. A 750w PSU is not pulling 750w from the wall at all times. It only pulls what the components need. 

 

Also, the majority of modern PSU's (anything 80+ bronze or better) is generally operating at near-peak efficiency across a large portion of it's load range (25-90% load).   

Agreed. On a single user cenario, it's irrelevant. But, multiply it by the thousands of cases where it happens, and some big numbers start to show up.

[MEC-777]

Agreed. On a single user cenario, it's irrelevant. But, multiply it by the thousands of cases where it happens, and some big numbers start to show up.

 

Actually I was referring to what that person said, whom I quoted, as being mostly irrelevant - that they see a lot of builds where the PSU is somewhat "overkill" for the system, implying that scenario would somehow use more energy than a system with a PSU that more closely matched the total system draw. If a PC only draws 350 watts, the PSU is only going to draw 350w (+ a few % relative to efficiency), whether it's a 550w PSU or a 1000w PSU.

[Counter-Strike Player]

If we don't eat them, they'll overpopulate and produce more greenhouse gasses. Unless you're suggesting we just slaughter all of them needlessly to protect ourselves.

 

Yes but you see, if we eat more, cattle farmers are going to recognize that as "increased demand" so they will attempt to breed more and hence more methane farts from the cows. So it's basically a lose-lose situation for humanity in general. gg.

[Trik'Stari]

Yes but you see, if we eat more, cattle farmers are going to recognize that as "increased demand" so they will attempt to breed more and hence more methane farts from the cows. So it's basically a lose-lose situation for humanity in general. gg.

Put a dome over the cattle ranchs, creates temporary jobs because someone has to build the domes, collect the methane from the cattle farts and use it as fuel.

 

Winning? There's nothing humanity can't overcome if we stop looking at things in terms of dollar cost. If anything will be our downfall, it will be our own adherence to an entirely fabricated and illogical system of currency.

[NatCat]

Well time is money to a lot of people.

Was simply replying to another poster who felt that the only thing to do was to shut down.  I agree that it's better to put effort in elsewhere, partly for the convenience factor you mentioned and partly because shutting down a suboptimized machine is kind of like turning a hummer off at red lights rather than rethinking what kind of car is needed to get the job done ... all assuming that gas prices or other side effects of energy use matter to the user.

[NatCat]

.... collect the methane from the cattle farts and use it as fuel.

 

Winning? There's nothing humanity can't overcome if we stop looking at things in terms of dollar cost. If anything will be our downfall, it will be our own adherence to an entirely fabricated and illogical system of currency.

 

And, believe it or not, even more methane comes from cow burps!   Third source is decomposition of manure.  Imagine the carbon footprint of a gamer cow.  Maybe we can all agree THAT's a problem

 

Completely agree with your more serious comment about strict short-term dollar costs.  That said, there will no doubt be some attractive ROIs on gaming PC energy improvements, but the mentality and data need to be available for people to even do the math correctly.

 

Thx.

[NatCat]

Don't worry, you're not the only one.

 

I think people get confused about this. "Power consumption" should not be a concern where performance is of utmost importance (when it comes to gaming), but "efficiency" should be a concern. More performance from the same or lower TDP GPU's is a very good thing, since it can lead to more overclocking headroom, less heat output, and the potential for even more powerful "bigger" GPUs. 

 

Unless someone has a legitimate concern with regards to power consumption (extreme high cost of local energy or small room that heats up easily, etc.), it should not be a deciding factor.

 

The point keeps being missed here that the study achieved as good or better performance with dramatically reduced energy use (and bills).  What is "utmost importance" is in the eye of the beholder.  Some people care about operating cost as well, especially mom and dad if they pay the power bill for their kids' gaming PC.  The typical gaming PC in the study uses about 1/3 as much electricity as the average California home -- i.e., not trivial.  The typical gaming PC costs from $140 to $560/year to run (depending on energy prices) - https://sites.google.com/site/greeningthebeast/cost-carbon - also not trivial, and this does not include what can be a comparable bump in air-conditioning use if you're in a hot climate (cold climate - sometimes this waste heat is useful throughout the house, but mostly not distributed well).

 

This can be a ton (literally) of CO2/year as well, which matters to some people.  A broader challenge will be that it matters very much to energy policymakers and regulaters at the city, state, national, and international levels.  Those folks routinely promulgate mandatory efficiency standards (think CAFE mpg standards for cars that have been on the books for decades now, without impeding performance, incidentally).  The regulators watch user communities to see if efficiency gains get made voluntarily.  If not, they often intervene.  Don't shoot the messenger (me) - I'm just pointing out how things might play out.

 

You're definitely right to flag "other legitimate concerns".  I think those are real here (even beyond broader energy/environmental considerations).  Efficient machines are going to be more reliable, quieter, etc.  Also, it's easy to oversize a PSU, which adds to first cost but not to performance.  Right-sizing is a win-win.  Same thinking goes with avoiding bottleneck situations (e.g., oversized GPUs).

 

[NatCat]

I think the consumer information out there is perfectly adequate .

 

Hard to agree here.  Many components' nominal power requirements are not labeled on the box, and can be hard to find even online. More importantly, the study shows that actual power draw is on average half of nameplate for complete system (varying by component, but almost always substantially less than nameplate). 

 

That's "good" when it comes to energy use, but, still in order to make informed decisions (and not oversize PSUs) we need better bench testing and labeling.  To do this meaningfully, there need to be industry standards and test procedures for this gear -- the study notes examples of how product reviews report significant variations in measured power demand for identical components and benchmarks.  Annual operating costs depend not only on the integral of (actual) energy use over all the modes of operation (see chart in study) but also on electricity tariff structures.  Electricity prices vary by a factor of 10 or more across the US today.  And, there are second-order effects like increased air-conditioning energy use that few people know how to estimate.  The information environment could be better.

[NatCat]

If people would only actually buy adequate PSUs, maybe that would help.

 

The amount of dudes I see with 750w PSUs, but only 300W-ish configs is insane....

 

MEC-777 was right that the PSU will only produce what is being called for by the downstream load.   As various folks noted, PSU efficiency varies as a function of load.  Fig 7 of the paper shows a few examples.  Oversizing the PSU does cause energy waste (lower efficiency at max load), but thankfully this effect has lessened in recent years.  It used to be that efficiencies droped to 20% and even lower at low part-loads.

 

Key consideration for money-minded people is that an oversized PSU costs more; wastes money (and may take up more room in the box, which isn't welcome).  Standby losses may be higher as well -- not sure about that.  So, one of the compound benefits of energy efficiency is helping to downsize (we like to call it right-size) the PSU.  It's a win-win as long as headroom is left for possible load expansion, but that shouldn't be an excuse for being lazy in spec'ing parts.

[mr moose]

The point keeps being missed here that the study achieved as good or better performance with dramatically reduced energy use (and bills).  What is "utmost importance" is in the eye of the beholder.  Some people care about operating cost as well, especially mom and dad if they pay the power bill for their kids' gaming PC.  The typical gaming PC in the study uses about 1/3 as much electricity as the average California home -- i.e., not trivial.  The typical gaming PC costs from $140 to $560/year to run (depending on energy prices) - https://sites.google.com/site/greeningthebeast/cost-carbon - also not trivial, and this does not include what can be a comparable bump in air-conditioning use if you're in a hot climate (cold climate - sometimes this waste heat is useful throughout the house, but mostly not distributed well).

 

This can be a ton (literally) of CO2/year as well, which matters to some people.  A broader challenge will be that it matters very much to energy policymakers and regulaters at the city, state, national, and international levels.  Those folks routinely promulgate mandatory efficiency standards (think CAFE mpg standards for cars that have been on the books for decades now, without impeding performance, incidentally).  The regulators watch user communities to see if efficiency gains get made voluntarily.  If not, they often intervene.  Don't shoot the messenger (me) - I'm just pointing out how things might play out.

 

You're definitely right to flag "other legitimate concerns".  I think those are real here (even beyond broader energy/environmental considerations).  Efficient machines are going to be more reliable, quieter, etc.  Also, it's easy to oversize a PSU, which adds to first cost but not to performance.  Right-sizing is a win-win.  Same thinking goes with avoiding bottleneck situations (e.g., oversized GPUs).

 

I disagree,  Introducing efficiency policy for any product will cause the cost of that product to go up.  Check the cost of cars in LA with the EPA laws, between $1000 and $3000 more per car.   The other side to the issue is that when efficient products saturate the market and the power companies aren't selling as much power resulting in reduced revenue, they will increase prices to counter that fall.  Thus there is no real saving for bill payers in the long run.  Significantly more power/money will be saved if people simply turn off their pc over night or while at school/work.

[NatCat]

I disagree,  Introducing efficiency policy for any product will cause the cost of that product to go up.  Check the cost of cars in LA with the EPA laws, between $1000 and $3000 more per car.   The other side to the issue is that when efficient products saturate the market and the power companies aren't selling as much power resulting in reduced revenue, they will increase prices to counter that fall.  Thus there is no real saving for bill payers in the long run.  Significantly more power/money will be saved if people simply turn off their pc over night or while at school/work.

 

"Efficiency policy" is a broad domain...

 

There are decades of history on this with utilities all over the country (and the world).  Utilities are actually spending $billions per year on energy efficiency programs and find it good for shareholders and consumers (lower bills -- it's the bill, not the per-kWh price that matters at the end of the day).  We encourage you to dig into it -- there is a huge literature beyond what can be summarized here.  More to the point, we're talking total cost of ownership.  The incremental cost of fuel-efficient cars pays back many times over in reduced fuel bills (not to mention public health and other externalities that aren't even priced into gas).

 

The other well-tested avenue is mandatory efficiency standards (utility programs are voluntary).  They too have shown to be highly cost-effective, saving $100s of $B for households via improved appliances, for example.  I'm just the messenger (don't shoot) -- mandatory standards will come if the regulators don't see the industry getting more on top of the energy use; even if a given gamer doesn't care, it imposes real costs (from energy infrastructure to public health) on everyone else.  Indeed, what increases people's energy bills is the energy inefficiency that drives the construction of new costly power plants.

 

Do you think 80Plus (PSUs) or EnergyStar (displays) have hurt the Gaming PC industry or imposed undue costs on gamers?

 

Good news is that your industry's own forward thinkers were onto this broader view five years ago now - http://jonpeddie.com/blogs/comments/calculating-the-total-cost-of-ownership1

 

Turning this gear off at night is like shutting down a hummer at red lights.  It saves some energy, but leaves much more savings on the table.

 

 

 

[mr moose]

 

Turning this gear off at night is like shutting down a hummer at red lights.  It saves some energy, but leaves much more savings on the table.

 

 

For an industry you think doesn't lose from efficiency programs they certainly are trying to minimize revenue loss. 

http://aceee.org/incentivizing-utility-led-efficiency-programs-program-cost-recovery-0

 

If you use your hummer 24/7 and don't need all the power of hummer then yes, but given most gamers don't game 24/7 and do actually use most of the performance of their systems, then absolutely turning off your pc at night will save more than making it more efficient. 

 

500watt system running 16 hours (all day and off for 8 hours over night)  =  8Kw/h

400Watt system (80% power usage)  running 24 hours = 9.6Kw/h

 

Simple turning the computer off for 8 hours a say and not leaving it running saves more power.

[ahhming]

A recently published study in the journal Energy Efficiency has found that gaming computers are using an incredible amount of energy. While only 2.5 percent of personal computers are classified as primary for gaming, they account for a staggering 20 percent of global computer energy use.
 

To put that into perspective, the study’s co-author Evan Mills offered this comparison, “it’s like 25 standard electric power plants,” or, “160 million refrigerators, globally.” While owning a beast gaming PC is awesome, gamers can find ways to make their machines more efficient. Mills calculates “a typical gaming computer uses 1,400 kilowatt-hours per year, or six times more energy than a typical PC and 10 times more than gaming console.”

The study finds the by replacing some components and tweaking settings, gamers could improve the performance and reduce their energy consumption by as much as 75 percent, amassing a potential global saving of over $18 billion by 2020, or enough to shut down 40 giant 500 megawatt coal plants.

 


 
 

“Your average gaming computer is like three refrigerators,” Mills said. “When we use a computer to look at our email or tend our Facebook pages, the processor isn’t working hard at all. But when you’re gaming, the processor is screaming."

Mills has also launched a site, Greening the Beast, to help gamers reduce their machine’s energy footprint, by providing tips for more energy efficient settings and kit. To strengthen their claim, Mills and his team have built their own hyper-efficient gaming rigs that can go toe to toe, benchmarked against the standard, energy sucking machines.

PSU (Seasonic G Series, 550 W), CPU (Intel Core i7 4820 K—quad core, 3.7 base GHz), GPU (NVIDIA Reference Geforce GTX 780, 900 MHz boost), motherboard (ASUS P9X79-EWS)

RAM (32GB (8×4 GB) Kingston HyperX Beast 1866 MHz, 1.65 V), display (Apple HD Cinema, 23 in.). Operating system: Windows 7 Professional 64 bit; BPower saver^ energy management settings in Windows7 OS.
 
By upgrading the above rig,the energy efficiency improvements from left to right, were upgraded to a 92% efficient PSU (Corsair AX760), improved GPU (Zotac Geforce GTX 970 AMP! Omega edition), improved motherboard (ASUS Sabertooth Z97 Mark I) and CPU (Intel Pentium G3258), and improved display (ASUS VG248QE modified with NVIDIA G-sync). Gaming performance remained essentially unchanged, resulting in nearly a doubling of system energy efficiency. Average machine has 1.4 GPUs, and thus higher energy use.

TLDR:
gaming PC  account for a  20 percent of global computer energy use, just by swapping some component, you'll get almost same performance lower  power draw
Source:
http://motherboard.vice.com/read/gaming-computers-use-a-truly-astonishing-amount-of-energy
 
https://docs.google.com/viewer?a=v&pid=sites&srcid=ZGVmYXVsdGRvbWFpbnxncmVlbmluZ3RoZWJlYXN0fGd4OjdhNTI3NzQ0MTJjYzY3MDE
 
http://newscenter.lbl.gov/2015/08/31/gaming-computers-offer-huge-untapped-energy-savings-potential/
 
https://sites.google.com/site/greeningthebeast/energy

[Paralectic]

lol....

[Khajiit Dealer]

lol....

 

This.

 

I pay for my electricity, I don't care how environmentally "unfriendly" it is. Hell, we're giving people at power plants jobs  :lol:

[Guest]

Like a man once said

With great power

Comes great bills to pay

[Kinda Bottlenecked]

Heh. Just made mah 390x use nearly 400w of power. 

[ADZ_123_!"£]

Like a man once said

With great power

Comes great bills to pay

[WolfDeville]

Heh. Just made mah 390x use nearly 400w of power. 

Shit!

 

Imagine how much it would use under load...

 

I kid

[soaringchicken]

Literally not surprising, like, at all.