Sign in with
Sign up | Sign in

Benchmark Results: Efficiency (Performance Per Watt)

Maximum Efficiency: Build A 25W Performance PC Using Core i5
By

Once again, we ran our efficiency suite, which consists of several scripted benchmarks.

Since we keep tracking power in one-second intervals, we can calculate the average power used for the entire workload. The difference is significant considering that the performance does not change. All we're doing is modifying components and voltage settings in an effort to decrease power consumption.

The results on total power used for this test run are similar to the average readings.

The total runtime is one form of a performance result. As you can see, the results are very similar for the first three configurations. The last test run, which was conducted with the slower but more efficient Toshiba 2.5” hard drive, delivers less overall performance due to the decreased disk performance.

We divided the performance result by the power used to get our efficiency score. As expected, efficiency increases quite a bit.

The efficiency diagram visualizes power consumption at every point of the benchmark run and also shows differences in runtime.

Display all 64 comments.
This thread is closed for comments
Top Comments
  • 11 Hide
    huron , February 23, 2010 10:02 AM
    abswindows7EVER THOUGHT OF A SSD ? SERIOUSLY ?



    From Step 3: An Efficient Hard Drive

    Of course, a solid state drive is also an option here, but the limited capacity would be an issue for most folks, and the increased cost would push the value of this build in the wrong direction. For that reason, a 2.5" mobile drive with more room for storage makes the best sense.


    It's amazing to me that people feel a need to yell (all CAPS) and have not actually read the article completely - maybe skimmed it and missed a section.

    I think the article is interesting, as that is a very low power goal. What was more surprising is how easy it was to accomplish.
Other Comments
  • 3 Hide
    noob2222 , February 23, 2010 5:40 AM
    Well, since my first comment was obviously deleted, I will try to make this one a little more meaningful.

    News Flash : This just in, we have nothing new to report so we will again try to show how Intel's dual core CPU rules.


    Just a suggestion, how about throwing together some useful articles on say cooling vs extremes. I would suggest something like this: setup 1, silent air cooling. setup 2, max air cooling, setup 3, silent water cooling, setup 4 max water cooling. Then put each through the paces, max OC, noise levels, and thermals.

    Personally all these i3 articles are leading up to the next SBM being yet another all Intel buildoff.
  • 2 Hide
    toms-viewer-1996 , February 23, 2010 5:42 AM
    Thank you for doing this review. I hate computer noise, and would like to build something as close to a fanless system as possible.

    I'm not sure how much the power supply in your review costs, but you may want to try the link below for a fanless, and high efficiency, 102W power supply. It cost about $80 with shipping. I think it averages about %80-90 efficiency at the 20-90 Watt level, and is completely silent.

    http://www.mini-box.com/picoPSU-150-XT-102-power-kit
  • 0 Hide
    hjk41 , February 23, 2010 6:30 AM
    toms-viewer-1996:
    I think we still need a fan to suck out the hot air produced in the PC case. So if the power supply does not have a fan, we will have to attach one to the case. However, it is not impossible to build a PC case which does not require a fan to suck out the hot air. I would be happy to see a product like that.


    I think it is strange that the article does not even mention replacing the hard disk with an SSD. Hard disks are still a major source for noise in modern PCs.

    I would be more than satisfied if they would test out some AMD platforms, such as Athlon X2 245+785G motherboard. I have a HTPC of this configuration. It is cool when idle, just that the power supply fan is too noisy. :-)
  • -8 Hide
    micky_lund , February 23, 2010 6:43 AM
    meh...shows intel raps up both ends of the spectrum: power and extreme high end, except for the price/performance parts (in low budget)
  • -8 Hide
    gladiator_mohaa , February 23, 2010 6:49 AM
    I do not even see the point of rating a computer based on efficiency. I mean understand that efficiency has its purpose for some products. Take cars for example, some people want to go fast and don't care about mpg and some do. It makes me laugh when I see when one HD is rate at 10 watts under load while the competition is 9 watts and the faster is uses 10 watts but is docked for that power consumption. Even if you ran a business that saving of a watt would probably cost you more when you figure the performance lost of your employees waiting for the HD to access. The initial price of the HD would matter more anyway because if the not so efficient HD is 30 dollars less, and efficient HD is 30 dollars more, what difference does it make that a hard drive will save 20 bucks over three years. You are still paying 10 bucks for "efficiency". Same goes for every computer product, just buy what you need, for the best price, at the best quality you can afford.
  • 2 Hide
    anamaniac , February 23, 2010 7:44 AM
    Cool.

    Get a high binned i5-660, and run as low a stable voltage as you possibly can with HT off (since atleast with the LGA 1366's, HT puts a heavy impact on the voltage needed for stability) with all luck, .20 vcore under, run dual channel DDR3 at 1066 and as low a voltage as possible ( I never noticed a difference between triple channel 1800MHz and dual channel 1066Hz on my i7 rig really), a 80GB Intel x18-m, which only cost $200 (network drive for everything else), and a 5970 with triple 30" 2560x1600 SPVA panels.
    =)
    Quote:
    I do not even see the point of rating a computer based on efficiency. I mean understand that efficiency has its purpose for some products. Take cars for example, some people want to go fast and don't care about mpg and some do. It makes me laugh when I see when one HD is rate at 10 watts under load while the competition is 9 watts and the faster is uses 10 watts but is docked for that power consumption. Even if you ran a business that saving of a watt would probably cost you more when you figure the performance lost of your employees waiting for the HD to access. The initial price of the HD would matter more anyway because if the not so efficient HD is 30 dollars less, and efficient HD is 30 dollars more, what difference does it make that a hard drive will save 20 bucks over three years. You are still paying 10 bucks for "efficiency". Same goes for every computer product, just buy what you need, for the best price, at the best quality you can afford.

    Hmm...
    1W/h*24h*365x3y= 26280W
    26.28KWh*$0.12/KWh= $3.15
    Assuming 24h/365d usage for 3 years, which I'm sure for most business computers, 8.5h a day is more likely, with 240 days a year (assuming 5 days a week, 20 days for vacation and sick days), that's 8.5h*240d*3y/1000*$0.12= $0.73

    I'm not going to give a bloody hell about 1 watt on a drones hard drive. I do care that my family members have an affliction with 100W incandescent light bulbs, as I'm more of a 15W fluorescent type myself.
  • 3 Hide
    HalfHuman , February 23, 2010 8:04 AM
    thank you for using a proper psu for doing efficency testing. you could have gone with something even smaller like ~100w. the hdd could have been easily replaced by a ssd not only for power usage but for performance, noise etc. a passive cooling solution could have dropped the power consumption even lower. guess the voltages could have been a lot lower, at least that is very possible on amd platforms. would have been nice to see a amd config built with efficency in mind (something like a power efficent 785 mobo and a dual core undervolted).

    i'm really happy to see a config built for efficency.
  • 1 Hide
    anonymous x , February 23, 2010 9:11 AM
    I like these builds. They are a refreshing break from the "oh lets built a bang for your buck gaming pc" that everyone does
  • -5 Hide
    Reynod , February 23, 2010 9:31 AM
    Yep ... another Intel paid for review by the guys who do it so well they will be forever remembered as the "IntelFanboitwins 1.0".

    Please rate me up and I will talk to the Nigerian Paymaster General on your behalf.

    This has been a short bout of unsponsored comic relief ...
  • -6 Hide
    Reynod , February 23, 2010 9:33 AM
    They blocked me from posting on any of their reviews ... I feel honoured.

    /AMD fans carry reynod off for victory lap ... unceremonially dumping him in uppity's trough.
  • -6 Hide
    Reynod , February 23, 2010 10:01 AM
    oops ...

    /facepalm
  • 11 Hide
    huron , February 23, 2010 10:02 AM
    abswindows7EVER THOUGHT OF A SSD ? SERIOUSLY ?



    From Step 3: An Efficient Hard Drive

    Of course, a solid state drive is also an option here, but the limited capacity would be an issue for most folks, and the increased cost would push the value of this build in the wrong direction. For that reason, a 2.5" mobile drive with more room for storage makes the best sense.


    It's amazing to me that people feel a need to yell (all CAPS) and have not actually read the article completely - maybe skimmed it and missed a section.

    I think the article is interesting, as that is a very low power goal. What was more surprising is how easy it was to accomplish.
  • -5 Hide
    notty22 , February 23, 2010 10:14 AM
    Great article. These cpu's possess features that are very important for today. Children will understand this when they have to pay the utility bills.
  • 2 Hide
    Anonymous , February 23, 2010 10:29 AM
    Anyone who has looked into low power computers knows that the PSU is the main thing to change to get efficiency. This is however always interesting reading to see just what figures you get with new hardware.

    It would have been really interesting to see comparisons with different CPUs like the cheaper i3-530 or to see how many W you would "sacrifice" to get up to a i5-750 and a cheap Ati 5xxx card.

    Nice work guys, I liked reading this article. =)
  • 1 Hide
    killerclick , February 23, 2010 11:16 AM
    If anyone thinks Tom's Hardware is pro-Intel, you should take a look at the Best Gaming CPU For The Money February article. No Intel CPUs are recommended below i5 750, it's all AMD.
  • 2 Hide
    FlorinR , February 23, 2010 11:21 AM
    Nice build. How much is it cost?
  • 0 Hide
    cknobman , February 23, 2010 12:10 PM
    Great article. As a system builder myself I love it when you do articles like this as it gives me new insights on my future builds. I always wonder about power draw and efficiency but since I do not have the proper equipment to measure it I rely on sites like Toms to do these articles.
  • -4 Hide
    obarthelemy , February 23, 2010 1:11 PM
    Useless article, reminiscent of those car oil consumption tests done at idle (in the US if memory serves right). If my PC is idle, is goes to suspend then sleep then hibernate farily quickly.

    I'd be interested in an "under typical conditions" test. That includes with a real HD, not one of those SSD geek toys.
  • 8 Hide
    noob2222 , February 23, 2010 1:25 PM
    killerclickIf anyone thinks Tom's Hardware is pro-Intel, you should take a look at the Best Gaming CPU For The Money February article. No Intel CPUs are recommended below i5 750, it's all AMD.

    Lets examine this. Starting with the i5 release date
    Jan 4- Clarkdale review
    Jan 4 - Mobile Clarkdale
    Jan 5 - Clarkdale efficiency vs AMD (ie 32nm vs 40nm)
    Jan 8 - P55 crossfire scaling
    Jan 12 - Efficiency Explored, whats the perfect clock rate for I5 clarkdale
    Jan 15 - Quad core cpus compared at 2.8ghz
    Jan 19 - cpu roundup
    Jan 25 - AMD PII 555 vs pentium 6950
    Feb 2 - AES-NI Performance Analyzed
    Feb 17 - Efficiency Explored - the perfect clock rate for the 15-661 (didn't we already see this?)
    Feb 22 - cpu roundup
    Feb 23 - Maximum efficiency- (yet another clarkdale article)

    So here we have 10 pro Intel CPU articles vs 1 AMD article and 2 cpu roundups ... Beating a dead horse anyone?
Display more comments