Page 1 of 2
1
2
LastLast
  1. #1
    Bloodsail Admiral Sturmbringe's Avatar
    Join Date
    Feb 2009
    Location
    Patras, Achaea, Greece
    Posts
    1,101

    What is your power consumption when running WoW at Ultra settings?

    Hello guys,

    I play WoW on ULTRA settings.

    My spec:

    CPU: Intel Yorkfield, Quad Core, Extreme QX 9650, Undercloked at 1.7/2.6 GHz.
    GPU: Gigabyte Nividia GTX 580, Underclocked at GPU clock 684 MHz and Memory Clock 820 Mhz.
    Memory: 4 Kingston 1 GB sticks, underclocked at 357.8 MHz.
    MOBO: ASUS P5Q Deluxe

    Power Consumption:


    CPU, idle: 6.79 Watts
    CPU, medium load: abt 12 Watts.
    CPU, playing WoW at Ultra settings : 15-25 Watts

    GPU, idle: 12.5 Watts
    GPU, medium load: abt 35 Watts
    GPU, playing WoW at Ultra settings :40-88 Watts (depending on game activity)
    =====================================================================

    I get abt 48-170 FPS, depending on location and how crowded the place is, or whether it's a raid or BG, etc.

    Edit: I play at 1920X1080, single monitor.
    Last edited by llDemonll; 2012-08-27 at 06:11 PM.

  2. #2
    I don't see what you are trying to accomplish. If you want to save power, get a low powered system. Thats like buying a large suv and driving it under the speed limit to save gas.
    :::: Intel 4770k w/ Corsair H80i ::::
    :::: 16gb Corsair Vengeance LP DDR3 1600 ::::
    :::: Corsair 550d Case ::::
    :::: EVGA GTX 780 SC ::::
    :::: Dell Ultrasharp U2713hm @ 2560x1440 ::::

    Pictures http://imgur.com/a/4JUcp

  3. #3
    Not exactly sure what you're measuring your consumption with, but your numbers are way, way off. A C2Q @ 2.6Ghz is still going to have a full load power consumption around 100w, and a GTX 580 underclocked by only 88Mhz is going to be around 220w under full load. Slap in around 35w for your motherboard, 8w per mechanical drive, 2w per fan (including CPU heatsink), and a few watts for your RAM and you're looking at around 400w under max load.

    Sure you're not maxing your hardware while playing WoW, but I have no doubt whatsoever that your setup is still pulling well over 250w playing it.

    If you want to get an accurate reading of your consumption, you need to get a power meter at your wall socket and take the power pulled and calculate the (advertised) efficiency of your PSU into that number.

    EDIT: Is it even worth your time to save the power? Standard rule of thumb in the US is that running a 100w light bulb 24/7 all month will cost around 10 bucks. Let's say your system at normal clocks pulls in 450w under full load. Now factor in that no game actually 100% maxes your system, and that you're obviously not gaming for 24 hours a day, 7 days a week. Idle (or even browsing / watching videos) computer systems pull tiny amounts of power around the 30w mark.

    So even if one gamed for 40 hours a week, you're looking at pulling 370w~ under gaming load for that duration. Divide that by the approximate 720 hours a month, and you get less power consumed than that 100w light bulb running 24/7 mentioned earlier.

    What does it all mean? Underclocking your system like that may save you a whopping $1.50 a month.
    Last edited by glo; 2012-08-27 at 06:37 PM.
    i7-4770k - GTX 780 Ti - 16GB DDR3 Ripjaws - (2) HyperX 120s / Vertex 3 120
    ASRock Extreme3 - JBL S300A - FiiO E10 - EVGA Supernova 650G - Corsair H80i

    build pics

  4. #4
    Your readings are completely inaccurate. No way that hardware is drawing so little, even if it is underclocked.

  5. #5
    Bloodsail Admiral Sturmbringe's Avatar
    Join Date
    Feb 2009
    Location
    Patras, Achaea, Greece
    Posts
    1,101
    Not exactly sure what you're measuring your consumption with, but your numbers are way, way off.
    No, they are not.

    My CPU readings are reported with ASUS EPU-6 Engine v 1.01.17.

    My GPU readings are reported through Gigabyte OC GURU Version 1.25.

    These readings are not quite as accurate as if I had a watt meter connected at the mains, but they are a quite accurate ballpark figure.

  6. #6
    Quote Originally Posted by Sturmbringe View Post
    No, they are not.

    My CPU readings are reported with ASUS EPU-6 Engine v 1.01.17.

    My GPU readings are reported through Gigabyte OC GURU Version 1.25.

    These readings are not quite as accurate as if I had a watt meter connected at the mains, but they are a quite accurate ballpark figure.
    Yeah, they are.

    http://www.pcper.com/reviews/Process...er-Consumption

    You're claiming your system consumes as much power while gaming as a C2Q + GTX 280 (lesser power consumption than a 580) setup does at idle.

    It's impossible. I'm not sure why you think that a 10% underclock on your CPU and GPU would result in a 60% reduction in power consumption. Simple logic applies here.
    i7-4770k - GTX 780 Ti - 16GB DDR3 Ripjaws - (2) HyperX 120s / Vertex 3 120
    ASRock Extreme3 - JBL S300A - FiiO E10 - EVGA Supernova 650G - Corsair H80i

    build pics

  7. #7
    /facedesk @ logic!

  8. #8
    Bloodsail Admiral Sturmbringe's Avatar
    Join Date
    Feb 2009
    Location
    Patras, Achaea, Greece
    Posts
    1,101
    No, they are not. I am not claiming anything, you are making your own claims you give your own responses. BTW, my CPU is a C2Q.

    Last edited by Sturmbringe; 2012-08-27 at 08:49 PM.

  9. #9
    Mechagnome
    Join Date
    Aug 2011
    Location
    Netherlands
    Posts
    617
    Did you just take a picture of your desktop? Teehee
    Last edited by protput; 2012-08-27 at 09:09 PM.
    CPU: I5 2500K @4.5ghz (1.285v), CPU cooler: Corsair H100 push/pull GPU: MSI R7970 Lightning @1200Mhz, Case: Corsair 650D, Ram: 16GB Corsair vengeance, Mobo: Asrock Z77 Extreme4, PSU: Corsair TX650M, SSD: Samsung 830 128GB, HDD: WD Caviar Black 1TB SPU: Xonar Essence STX

  10. #10
    Titan Synthaxx's Avatar
    Join Date
    Feb 2008
    Location
    Rotherham, England/UK
    Posts
    13,103
    Here's the setup with a ;
    Intel Core i7-965 processor (3.2GHz: 133MHz x 24)
    Asus P6T V2 motherboard (Intel X58 Express with three PCI-Express 2.0 x16 slots)
    3x 2GB Corsair TR3X6G1333C9 memory modules (operating in dual channel at DDR3 1,600MHz 9-9-9-24-1T)
    Corsair X128 120GB SSD running v1 firmware
    Corsair HX1000W PSU
    Windows 7 Home Premium x64
    At idle, the GTX580 system uses 150W. At load, in 3DM06 Canyon test, it bolts up to 363W (http://www.bit-tech.net/hardware/gra...x-580-review/8).

    Thus, we can assume that the mean usage of the GPU alone will be approx. 213W when under heavy load. A single GTX580 at full ultra in WoW at 1080P sits at around 75% usage, though we have to discount some of that for parts of the GPU that aren't working flat out at 75% and thus it's safe to assume it's 55% "true usage". This brings us around 117W usage. If we really wanted to push it, 90W would be about the minimum.

    The QX9650 setup was;
    Intel Core 2 Duo QX9650
    Gigabyte X38DQ6
    Corsair PC2-8500 4GB
    XFX GeForce 8800 Ultra
    Western Digital SATA 250Gb
    Corsair Nautilus 500
    Corsair HX620W
    Windows Vista Ultimate
    The QX9650 in that setup reportedly used 168 at idle, and 222 at load using POV-Ray meaning the CPU itself uses approx. 54W under load. WoW uses approx 40% usage on a quad core CPU (though this can vary based on the CPU) and thus we can assume around 21W usage for the CPU (http://www.legitreviews.com/article/583/11/).

    We're talking 138W there as a rough guess as usage for the CPU and GPU alone. However, none of us are right to say what's true and what isn't, even the OP, because it's much more valid to measure at the socket values for the whole system than making vague assumptions, and more importantly, trusting a piece of software to measure something like this.
    Ex-coder, gamer - Stepped away from dev.
    Done/Planned: Leedsfest 2014, Mallory Knox (Nov 2014), Deaf Havana (Dec 2014) | Upcoming: Soundwave (Feb 2015)
    Awaiting UK/EU tour dates: Of Mice and Men, Jimmy Eat World, Architects (>Mar. 2015)
    Quote Originally Posted by George View Post
    If NSA is building my profile from tonight's "browsing habits", they're in for a treat: women with three breasts, men with two penises, fake nipples...

  11. #11
    Bloodsail Admiral Libram's Avatar
    Join Date
    Nov 2011
    Location
    Delta Quadrant
    Posts
    1,191
    Quote Originally Posted by protput View Post
    Did you just take a picture of your desktop? Teehee
    Made me laugh a bit to xP
    I make Trance, let me know what you think! SoundCloud YouTube


  12. #12
    Bloodsail Admiral Sturmbringe's Avatar
    Join Date
    Feb 2009
    Location
    Patras, Achaea, Greece
    Posts
    1,101
    @Synthaxx:

    21W usage for my CPU at stock clock speeds sounds about right. I got about 25 W reading with ASUS EPU Engine with my CPU underclocked at 2.6 GHz.

    Also, my CPU is a Core 2 Quad not a Core 2 Duo.

  13. #13
    Mechagnome
    Join Date
    Aug 2011
    Location
    Netherlands
    Posts
    617
    So you're saying a 125 watt power supply would be enough for your build?

    Find a 125 watt PSU that you can use in your build, I dare you.
    CPU: I5 2500K @4.5ghz (1.285v), CPU cooler: Corsair H100 push/pull GPU: MSI R7970 Lightning @1200Mhz, Case: Corsair 650D, Ram: 16GB Corsair vengeance, Mobo: Asrock Z77 Extreme4, PSU: Corsair TX650M, SSD: Samsung 830 128GB, HDD: WD Caviar Black 1TB SPU: Xonar Essence STX

  14. #14
    I am Murloc! Cyanotical's Avatar
    Join Date
    Feb 2011
    Location
    Colorado
    Posts
    5,204
    underclocking is pointless, CPUs and GPUs throttle when idle to save power, nor does it let you run at the same performance as stock, it gives almost no advantage anywhere, your games run slower, and you save diddly on your power bill, nor does it save the environment, computers are nothing more than organized functional pieces of toxic waste

    also, software doesn't show true power draw

    i7-4790K | Z97 Class. | 8GB DDR3-2133 | GTX-690 Quad SLI | RAIDR | 512GB Samsung 830 | AX1200 | RV05
    Dell U2711 | Ducky Shine3 YoS | Steelseries Sensei | Xonar Essence One | KRK RP8G2s

  15. #15
    Titan Synthaxx's Avatar
    Join Date
    Feb 2008
    Location
    Rotherham, England/UK
    Posts
    13,103
    Quote Originally Posted by Sturmbringe View Post
    Also, my CPU is a Core 2 Quad not a Core 2 Duo.
    AFAIK there's only 1 QX9650. There's not a "C2Q QX9650" or a "C2D QX9650", it's just QX9650 so it's the same end result either way (i did just quote the specs from those sites and it's been 3-4 years since i dealt with a Core 2 chip). I also realised i made a bit of a mistake. The 138W value i gave was the INCREASE over idle, since i couldn't ascertain idle power usage from the reviews alone so the actual value would be 138W + Idle wattage.

    As others have said, underclocking is pointless. Power not used it power not drawn. This is why measurements "at the socket" in many reviews are often WAY below what a PSU is capable of. A PSU only pulls from the wall what is being used by the components. If the system is using 50W, then only 50W is drawn from the wall. If the system is using 1200W, then 1200W is drawn from the wall.

    Underclocking will drop power usage, yes, but realistically, you're only really going to be drawing 550W max with that system. That means, every 24 hours the system is running, it's using 13200 kWh. On a realistic 8 hour session, we're talking 4400 kWh. Now, maybe those values mean nothing to you. But, let's underclock and say we save 50W (that's a VERY generous deduction too). We're now at 4000 kWh for that 8 hours. I'll refer to 8 hour sessions from now on as "a day" or "per day". At 4400 kWh, it's costing £0.38 per day. At 4000 kWh, it's costing £0.35 per day (theoretical values). Over the course of a year, you save £12.70. Now, sure, any money saved is good, but these are such miniscule amounts that the performance you lose from gaming and other tasks just isn't worth it in most cases.

    To add in another point, i once saw my 2600K, with hyperthreading enabled, "downclock" to 1.6Ghz and according to HWMonitor, use 6.2W of power. We're talking a 4-core hyperthreaded chip using a ridiculously low amount of power. My GTX580's clocked down to 2D speeds too, and i got 18W on each card (3 cards total). Despite this illusion of low power usage, i left it on overnight after forcibly setting the CPU to run at 1.6Ghz (by building a piece of software to limit the Windows CPU Power usage to 0%). In the morning (10 hours later), it'd used up £0.44. I then set it to not downclock when idle (and overclocked it to 4.6Ghz) and the 580's to stay at 3D performance clock values, and left it again for 10 hours the next night. In the morning, it'd used up £0.49. In both cases, the system was essentially idle.

    There are 8765 hours in a year. At default of £0.49 per 10 hours, it costs £429 to run it for a year. At £0.44, it costs £385 to run it for a year. Of course, this last part seems a little offtopic, but can you imagine playing games, or video rendering, or doing anything similar when you're getting 65% less performance on the CPU, and 93% less performance on the GPU's? Now can you really see it being worth saving £44 per year while getting an average of 79% less performance overall? What used to run smoothly at 60FPS is now running at 12.6 FPS. A video render that used to take 30 minutes now takes almost 54 minutes. Bear in mind, those savings are at idle, but the percentage drawn by moving up to load in both cases will be identical.

    This is almost all theoretical, but the point i'm trying to make is that underclocking rarely saves you much cash, especially when you consider the extra time it takes to complete things, the lower performance, and the potential stress this could cause.
    Ex-coder, gamer - Stepped away from dev.
    Done/Planned: Leedsfest 2014, Mallory Knox (Nov 2014), Deaf Havana (Dec 2014) | Upcoming: Soundwave (Feb 2015)
    Awaiting UK/EU tour dates: Of Mice and Men, Jimmy Eat World, Architects (>Mar. 2015)
    Quote Originally Posted by George View Post
    If NSA is building my profile from tonight's "browsing habits", they're in for a treat: women with three breasts, men with two penises, fake nipples...

  16. #16
    Bloodsail Admiral Sturmbringe's Avatar
    Join Date
    Feb 2009
    Location
    Patras, Achaea, Greece
    Posts
    1,101
    @Synthaxx:

    Core 2 Duo were the earliest Core 2's made, and as the name suggests, they were a dual core (2 cores) achitecture CPU.

    Core 2 Quads are the latest Core 2's made, and as the name suggests, they are a four-core chip (2x Core 2 Duos per).

    ----------------------->So, there isn't such a thing as a Core 2 Duo QX 9650, but there actually is such a thing as a C2Q QX 9650.

    the extra time it takes to complete things, the lower performance, and the potential stress this could cause.
    I am not seeing any of these. WoW has got the same performance for me with my CPU OC'd @ 4.2 GHz as it has with my CPU UNDERclocked at 2.66 GHz. This is actually the reason why I underclocked it in the first place.

    ^^From my experience as noted above, I conclude that MY CPU's speed has got absolutely NO effect whatsoever on WoW. The GPU does, however, have a rather huge effect on WoW performance. The performance difference between my old ATI RADEON HD 5830 and my new Nividia GTX 580 is rather huge.

    ----------------------->In addition, operating a CPU and a GPU underclocked, means that both parts work at lower temperatures, thereby REDUCING the temperature induced wear-and-tear both on the said components as well as on the MOBO chipset itself.

    Further, there is no such thing as "stress" applicable to an $1k Extreme CPU such as mine, especially when it works underclocked.
    Last edited by Sturmbringe; 2012-08-28 at 01:19 PM.

  17. #17
    Quote Originally Posted by Sturmbringe View Post
    @Synthaxx:

    Core 2 Duo were the earliest Core 2's made, and as the name suggests, they were a dual core (2 cores) achitecture CPU.

    Core 2 Quads are the latest Core 2's made, and as the name suggests, they are a four-core chip (2x Core 2 Duos per).

    ----------------------->So, there isn't such a thing as a Core 2 Duo QX 9650, but there actually is such a thing as a C2Q QX 9650.



    I am not seeing any of these. WoW has got the same performance for me with my CPU OC'd @ 4.2 GHz as it has with my CPU UNDERclocked at 2.66 GHz. This is actually the reason why I underclocked it in the first place.

    ^^From my experience as noted above, I conclude that MY CPU's speed has got absolutely NO effect whatsoever on WoW. The GPU does, however, have a rather huge effect on WoW performance. The performance difference between my old ATI RADEON HD 5830 and my new Nividia GTX 580 is rather huge.

    ----------------------->In addition, operating a CPU and a GPU underclocked, means that both parts work at lower temperatures, thereby REDUCING the temperature induced wear-and-tear both on the said components as well as on the MOBO chipset itself.

    Further, there is no such thing as "stress" applicable to an $1k Extreme CPU such as mine, especially when it works underclocked.
    I have no more faith in this topic, your not pulling less than mobile wattage from your desktop, if you think you are then your ignorant to all the facts put in front of you and argumentative for nothing. On topic I use nothing green, I drive a big ass cadillac escalade, I love big macs in the box and sometimes I just spray aerosol into the ozone. Guess my wattage @ peak.

    Also reducing the speed of your cpu makes it harder for the CPU to do things, i.e STRESS not the same thing as doing a stress test to induce a bsod.

    ---------- Post added 2012-08-28 at 03:25 PM ----------



    This just proves a number of things, closest frequency shows 6watts @ 2.4Ghz we will chock that up to voltage.
    However the chances your voltage is low enough to half the wattage @ 2.4Ghz is highly unlikely.

    Furthermore this will show likely a semi overclocked system of a much more efficient design @ idle with a 580 and many other gpus'
    Last edited by Milkshake86; 2012-08-28 at 03:30 PM.

  18. #18
    Bloodsail Admiral Sturmbringe's Avatar
    Join Date
    Feb 2009
    Location
    Patras, Achaea, Greece
    Posts
    1,101
    @Milkshake

    Thanks for taking the time to find those great tables. I maintain that an Extreme CPU like mine cannot be considered stressed anywhere below 4.0 Ghz. And, oh yeah, I love green.

  19. #19
    Quote Originally Posted by Sturmbringe View Post
    @Milkshake

    Thanks for taking the time to find those great tables. I maintain that an Extreme CPU like mine cannot be considered stressed anywhere below 4.0 Ghz. And, oh yeah, I love green.
    Your missing the part that overclocking doesnt STRESS your cpu, it alleviates stress, requiring more to bog it down. So your actually increasing the stress to your system by underclocking. Again we are not talking about wear and tear from higher temps.

  20. #20
    Titan Synthaxx's Avatar
    Join Date
    Feb 2008
    Location
    Rotherham, England/UK
    Posts
    13,103
    Quote Originally Posted by Sturmbringe View Post
    Further, there is no such thing as "stress" applicable to an $1k Extreme CPU such as mine, especially when it works underclocked.
    I'm sorry, but if you believe this, then i have to call you out on it. There are applications that will stress a CPU regardless of it's model. It was an extreme model... 4.5 years ago. It does not matter what it cost.

    Prime95 with 4 threads and heavy FFT will stress ANY 4-thread CPU. You can put a 16 core Opteron in there (yes, they do exist, have done since November 2011), set P95 for 16 threads with heavy FFT and it will indicate 100% usage on every single core. No exceptions. Stress tests care not for cost, nor for model or brand or release date. They care only for eating up as much CPU time as possible.

    To put your statement into perspective, your QX9650 that cost $1K in late 2007, is now worse than my 2600K that cost £250 (which equates to around $250 USD due to UK tax/VAT) from January last year (http://www.xtremesystems.org/forums/...=1#post4711835).

    I'm sorry, and it's nothing personal, but i absolutely hate when we get someone "new" to the Computer section of these forums tries to go against what the regulars know to be true, only for the sake of convincing themselves that their 4 year old system is still the new hotness. We've had someone come in before proclaiming 60FPS at Ultra in 25-man Firelands from an Athlon 64 Dual core (and not one of the later models, we're talking a 3500+), 2GB RAM, ATI 3850, and Windows XP. They gave us 2 screenshots; 1 of their settings, another of an in-game screenshot.

    The settings screenshot showed all max settings, with them looking at the ground in Orgrimmar. The other screenshot was from Lord Rhyolith in Firelands. Sure enough, they'd gotten everything set to low to try and prove to us their system was pumping out more FPS than systems built this year from the top end components. To explain the second screenshot, from the edge of Rhyolith's platform, the "view distance fog" was already starting to show in front of the boss, and the texture quality evidently wasn't ultra. Shadows were the low "blob" quality.

    Essentially, what i'm trying to say is that you shouldn't ever try to bullshit people on this forum. Almost all of the regulars have seen it all before and eventually it just becomes a tiresome game. Again, it's nothing personal, but we've seen all these proclamations from people who want us to believe their system is still awesome because they paid $1K for the CPU 4 years ago before.
    Ex-coder, gamer - Stepped away from dev.
    Done/Planned: Leedsfest 2014, Mallory Knox (Nov 2014), Deaf Havana (Dec 2014) | Upcoming: Soundwave (Feb 2015)
    Awaiting UK/EU tour dates: Of Mice and Men, Jimmy Eat World, Architects (>Mar. 2015)
    Quote Originally Posted by George View Post
    If NSA is building my profile from tonight's "browsing habits", they're in for a treat: women with three breasts, men with two penises, fake nipples...

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •