Please report all spam threads, posts and suspicious members. We receive spam notifications and will take immediate action!
Page 1 of 4 123 ... LastLast
Results 1 to 10 of 37

Thread: ASRock H77 Pro4-M - unreasonable high power consumption?




  1. #1
    Join Date
    Feb 2013
    Location
    Germany
    Posts
    8

    Default ASRock H77 Pro4-M - unreasonable high power consumption?

    To use it as a home file server, I bought the following system
    CPU: Intel Celeron Dual-Core G1610
    Mainboard: ASRock H77 Pro4-M, H77
    RAM: 2x 4 Gb
    HDD: Seagate Barracuda Green 5900.3 2000GB, SATA 6Gb/s
    PSU: Tagan 2-Force II 500W ATX 2.2
    Cooler: Xigmatek Loki SD963
    OS: Debian Sqeeze

    The G1610 paired with a reasonable mainboard and PSU should have a idle power consumption of about ~20 Watt. My system sucks 32 Watt on idle, 51 Watt on load. CPU, HDD and PSU have been well tested for low power consumption, so I wonder where all my energy goes. Did anybody already measure his idle power consumption with the ASRock H77 Pro4-M and can verify my observations?

  2. #2
    parsec's Avatar
    parsec is offline Senior Member
    Join Date
    Jun 2012
    Location
    Third stone from the sun
    Posts
    4,824

    Default Re: ASRock H77 Pro4-M - unreasonable high power consumption?

    What are you using to measure power consumption? Correct readings require an accurate measuring device, which are not cheap.

    I imagine you have all the CPU power saving features enabled, like EIST (Speed Step), C1E, or other C States if your CPU supports them. My ASR board has a power saving feature in the AXTU utility, called IES, that must be enabled to be active.

    CPU cooler and PC case fans easily use 1 Watt of power each or more, regardless of their size.

    Your PSU seems to not be a recent model, first introduced in 2007 - 2008. The data sheet I found for it stated 80% efficiency, but no mention of the 80 Plus certification for PC PSU's. 80 Plus certification existed when this unit was introduced, but I couldn't find any certification for it. An 80% efficiency figure for this unit is reasonable, and may be generous.

    Which means, 100 Watts in to the PSU, 80 Watts output power to the PC. 50 Watts in, 40 Watts out. 32 Watts in, produces 25.6 Watts output to the PC. Is this accounted for in your power usage measurement or calculation?

    But the efficiency specification is not that simple, it varies depending upon the amount of power the PSU is producing. The 80 Plus certification starts at a 20% of the rated power load on the PSU. It turns out that PC PSU's are more efficient at higher output than at low output. In detailed PSU testing as done at jonnyGURU.com, it is a given for a PSU at a 10% load to have lower efficiency than at 20% or 50% loads. Your PSU at an actual 20 Watt load, which is less than 5% of its capacity, will be under 80% efficiency, hopefully ~75%, and likely lower. Brand new 80 Plus Gold and Platinum PSUs that are over 90% efficient at 20% and 50% loads, are ~80% efficient at a load of 10%.

    Your goal of a very low power usage PC illustrates one of the problems in reaching that goal, the reduced efficiency of PC power supplies at very low loads. At some point reducing the load on the PSU will not lower the overall power used, as the PSU's efficiency drops off, and nothing or very little is gained.

    If you are really using only 50 Watts at load, you're doing very well.

  3. #3
    Join Date
    Feb 2013
    Location
    Germany
    Posts
    8

    Question Re: ASRock H77 Pro4-M - unreasonable high power consumption?

    Quote Originally Posted by parsec View Post
    What are you using to measure power consumption? Correct readings require an accurate measuring device, which are not cheap.
    That's true. I've got a cheap measuring device. I've validated the error by measuring 25W, 40W light bulbs to be <10% at this range.

    Quote Originally Posted by parsec View Post
    CPU cooler and PC case fans easily use 1 Watt of power each or more, regardless of their size.
    Unplugging all fans except the CPU cooler gained 3W.

    Quote Originally Posted by parsec View Post
    Your PSU seems to not be a recent model, first introduced in 2007 - 2008. (...)
    Your goal of a very low power usage PC illustrates one of the problems in reaching that goal, the reduced efficiency of PC power supplies at very low loads.
    I'm aware of the PSU problem. It was a spare PSU and 10% extra loss at 20W over 6 years isn't worth buying a new PSU. BUT you're completely right to mention that the efficiency of the PSU is first guaranteed from 20% load. Still, I wonder if it's so bad, most other servers /low power PCs would have the same problem.

    Quote Originally Posted by parsec View Post
    I imagine you have all the CPU power saving features enabled, like EIST (Speed Step), C1E, or other C States if your CPU supports them. My ASR board has a power saving feature in the AXTU utility, called IES, that must be enabled to be active.
    Yes, C1E, C3, C6 are on in the BIOS. I already asked the support to clarify the "Package C State Support", "Render Standby" and "Deep Render Standby" CPU / GPU options. All unnecessary ports except USB and Video are off, as well as all hot plugging.

    IES is a mystery to me, because I use Linux Debian Squeeze. What does IES stand for? How does it work? Can I enable it using Linux?

    I also noticed just a minor power consumption drop between bios and idle. Bios is 40W, idle is 32W. My main HDD doesn't spin down yet, probably because of logging.

    Quote Originally Posted by parsec View Post
    If you are really using only 50 Watts at load, you're doing very well.
    Undervolting the CPU by 0.15V to 0.95V brings me to 46W under full load (prime95). Yet, load power consumption isn't the issue as the home server will idle most of the time. Powertop shows the CPU uses most of it's potential, so I'm still wondering if the ASRock H77 Pro4-M has an unreasonable high idle power consumption.

  4. #4
    Join Date
    Dec 2009
    Location
    Austin TX
    Posts
    10

    Default Re: ASRock H77 Pro4-M - unreasonable high power consumption?

    Quote Originally Posted by parsec View Post
    Your goal of a very low power usage PC illustrates one of the problems in reaching that goal, the reduced efficiency of PC power supplies at very low loads. At some point reducing the load on the PSU will not lower the overall power used, as the PSU's efficiency drops off, and nothing or very little is gained.
    Agreed - matching PSU to system is really where the efficiency comes in. (See Tom's article on max efficiency)

    However, to the original poster, a conventional PSU may be your problem. I have 2 low power systems that follow the same issue you had. Using kill-a-watt (calibrated as reasonably accurate) *** only the system plugged in to it.

    AMD A4-3300 (32nm cpu) with MSI A75A-G55 mobo... sata hdd and DVD one case fan with an Antec 380 (80+ bronze) drew about 31 Watts

    I then connected a 65W PSU from an Antec ISK 300-65 (basically a laptop powerbrick)

    consumption went down to 22W

    Just did a setup with an MSI H61M-P31 and intel G1610 celeron (22nm) using an antec 500W (it was convenient on my bench) with sata hdd and DVD drive drew about 28W

    With the 65W antec the draw was 19-21 W (basic desktop use)

    This does not mean small PSU's are more efficient (often they are not -- Johnny Guru has some interesting reviews on Pico PSU's here) but matching the PSU output to the system max leads to lower overall consumption (which is the bottom line)
    Unfortunately, like Detroit selling SUV's, PSU makers love to sell bigger rigs since most people wont pay for high efficiency low output.
    Last edited by AcmeGadgets; 02-22-2013 at 12:48 AM.

  5. #5
    profJim's Avatar
    profJim is online now Chief Munchkin + moderator
    Join Date
    Dec 2008
    Location
    Tacoma, WA. [USA]
    Posts
    7,811

    Question Re: ASRock H77 Pro4-M - unreasonable high power consumption?

    Sounds like good advice but two questions come to mind:
    1. Can a 65 watt powerbrick reliably handle 24/7 continuous operation?
    2. How good is the powerbricks voltage regulation and ripple control?
    Q9650 @ 4.10GHz [9x456MHz]
    P35-DS4 [rev: 2.0] ~ Bios: F14
    4x2GB OCZ Reaper PC2-8500 1094MHz @5-5-5-15
    MSI N460GTX Hawk Talon Attack (1GB) video card <---- SLI ---->
    Seasonic SS-660XP2 80 Plus Platinum psu (660w)
    WD Caviar Black WD6401AALS 640GB (data)
    Samsung 840 Pro 256GB SSD (boot)
    SLI @ 16/4 works when running HyperSLI
    Cooler Master 120XL Seidon push/pull AIO cpu water cooling
    Cooler Master HAF XB computer case (RC-902XB-KKN1)
    Asus VH242H 24" monitor [1920x1080]
    MSI N460GTX Hawk (1GB) video card
    Logitech Z-5500 Digital 5.1 Speakers
    win7 x64 sp1 Home Premium
    HT|Omega Claro plus+ sound card
    CyberPower CP1500PFCLCD UPS
    E6300 (R0) @ 3.504GHz [8x438MHz] ~~ P35-DS3L [rev: 1.0] ~ Bios: F9 ~~ 4x2GB Kingston HyperX T1 PC2-8500, 876MHz @4-4-4-10
    Seasonic X650 80+ gold psu (650w) ~~ Xigmatek Balder HDT 1283 cpu cooler ~~ Cooler Master CM 690 case (RC-690-KKN1-GP)
    Samsung 830 128GB SSD MZ-7PC128B/WW (boot) ~~ WD Caviar Black WD6401AALS 640GB (data) ~~ ZM-MFC2 fan controller
    HT|Omega Striker 7.1 sound card ~~ Asus VH242H monitor [1920x1080] ~~ Logitech Z-5500 Digital 5.1 Speakers
    win7 x64 sp1 Home Premium ~~ CyberPower CP1500PFCLCD U.P.S
    .


  6. #6
    Join Date
    Dec 2009
    Location
    Austin TX
    Posts
    10

    Default Re: ASRock H77 Pro4-M - unreasonable high power consumption?

    Quote Originally Posted by profJim View Post
    Sounds like good advice but two questions come to mind:
    1. Can a 65 watt powerbrick reliably handle 24/7 continuous operation?
    2. How good is the powerbricks voltage regulation and ripple control?
    Don't know for sure (but won't let that stop me...) The Johnny Guru article addresses some ripple issues, some bricks better than others -- laptops tend to be more tolerant of variability.

    As to handling 24/7 continuous operation, it's quite possible. As a former IBM employee I had several laptops in docking stations as my desktop machines. They would be plugged in 24/7 (often without sleep or hibernate mode) for months at a time. Never had a psu related failure (nor a spontaneous laptop fire for that matter... tho' that is usually the battery as Boeing now learns :-).

    I may have to run some server experiments.
    (The real advantage of the Antec ISK case was the converter takes the power brick input and has plugs for the mainboard and sata connectors... but I wish Antec sold the converter only since I've got several laptop power units I can use)

  7. #7
    parsec's Avatar
    parsec is offline Senior Member
    Join Date
    Jun 2012
    Location
    Third stone from the sun
    Posts
    4,824

    Default Re: ASRock H77 Pro4-M - unreasonable high power consumption?

    AcmeGadgets, you've done exactly what I wondered about, using a small PSU that (hopefully) is more efficient at low loads. I'm glad to see it works, but finding a PSU like that needs research. As you said, the focus of PSU manufactures is more towards high power, the kilowatt and beyond units for multi-video card use. The lower power, fanless PSUs like Seasonic has, a 400 Watt model seems like it might have better efficiency at low loads, but those units are basically their ~600 Watt models without fans. Checking the JGuru review on it will show the reality of its low load efficiency.

    Stephanq, Package C State Support are additional CPU power saving options that will definitely reduce CPU power usage. You should enable all the C State options, from C1E, to C6, which is the maximum power saving setting. C6 will literally stop a core from executing instructions, the core clock is literally shut off/gated to a core. As soon as any CPU load is present, the core is activated, so these options are great for systems that may be idle for long periods.

    The Render Standby and Deep Render Standby options I am not 100% clear on how they work. They are worth using.

    IES is an option in the AXTU utility, which runs in Windows, but I don't know if it will work with your OS. My board does not have a BIOS setting for it, so I doubt others do. It stands for Intelligent Energy Saver, but the info about it on ASR's site only mentions AMD Phenom boards, which may be inaccurate, since it is available on my Intel system. It reduces the power usage of the CPU voltage regulator modules (VRM), apparently by turning off some of the VRM chips.

    Power usage in the BIOS is higher because apparently some or all of the CPU power saving options are not active at that point.

    Regarding a mother boards power usage, I have never seen that tested. The amount of information on that is very small AFAIK. Your board has 1/2 the VRM chips my board does for the CPU, but it's not that simple. It depends on the chips, are they efficient or not? Cheaper boards tend to use cheaper parts, so they may not be the best chips efficiency wise. I have seen users having problems when they OC the CPU on your board, it seems to run out of CPU power. So the power usage of a board is basically a mystery, IMO.

    If you don't use any of the following, you can disable them: onboard audio, the extra ASMedia SATA chip, CPU virutalization.

  8. #8
    profJim's Avatar
    profJim is online now Chief Munchkin + moderator
    Join Date
    Dec 2008
    Location
    Tacoma, WA. [USA]
    Posts
    7,811

    Default Re: ASRock H77 Pro4-M - unreasonable high power consumption?

    I need to add a 3rd question:
    Sounds like good advice but two questions come to mind:
    1. Can a 65 watt powerbrick reliably handle 24/7 continuous operation?
    2. How good is the powerbricks voltage regulation and ripple control?
    3. What is the maximum wattage that your system might need?

    Whatever the answer is to question 3, you'll need to add a safety fudge factor of 30 - 40% for your powerbrick or psu.
    Q9650 @ 4.10GHz [9x456MHz]
    P35-DS4 [rev: 2.0] ~ Bios: F14
    4x2GB OCZ Reaper PC2-8500 1094MHz @5-5-5-15
    MSI N460GTX Hawk Talon Attack (1GB) video card <---- SLI ---->
    Seasonic SS-660XP2 80 Plus Platinum psu (660w)
    WD Caviar Black WD6401AALS 640GB (data)
    Samsung 840 Pro 256GB SSD (boot)
    SLI @ 16/4 works when running HyperSLI
    Cooler Master 120XL Seidon push/pull AIO cpu water cooling
    Cooler Master HAF XB computer case (RC-902XB-KKN1)
    Asus VH242H 24" monitor [1920x1080]
    MSI N460GTX Hawk (1GB) video card
    Logitech Z-5500 Digital 5.1 Speakers
    win7 x64 sp1 Home Premium
    HT|Omega Claro plus+ sound card
    CyberPower CP1500PFCLCD UPS
    E6300 (R0) @ 3.504GHz [8x438MHz] ~~ P35-DS3L [rev: 1.0] ~ Bios: F9 ~~ 4x2GB Kingston HyperX T1 PC2-8500, 876MHz @4-4-4-10
    Seasonic X650 80+ gold psu (650w) ~~ Xigmatek Balder HDT 1283 cpu cooler ~~ Cooler Master CM 690 case (RC-690-KKN1-GP)
    Samsung 830 128GB SSD MZ-7PC128B/WW (boot) ~~ WD Caviar Black WD6401AALS 640GB (data) ~~ ZM-MFC2 fan controller
    HT|Omega Striker 7.1 sound card ~~ Asus VH242H monitor [1920x1080] ~~ Logitech Z-5500 Digital 5.1 Speakers
    win7 x64 sp1 Home Premium ~~ CyberPower CP1500PFCLCD U.P.S
    .


  9. #9
    Join Date
    Dec 2009
    Location
    Austin TX
    Posts
    10

    Default Re: ASRock H77 Pro4-M - unreasonable high power consumption?

    I agree with Parsec's observations about settings and bios power (bios is designed to run with no frills). Finding low output PSU's is the challenge. But to Stefanqn's original question it does seem like a problem... using the G1610 with his configuration even with a 500W PSU should not be using more than 30-40W. (I'm using the stock heatsink which is minimal, not sure how much the fan on the Xigmatek draws?)

    Stefanqn -- how old is the Tagan? (or how much power does it draw when idle (that is, when the PSU is ON but the system is OFF) Most PSU's will draw between 2-5W with the PSU rocker switch on. Older or inefficient PSU's can draw quite a bit more... I have some that will draw 12W when plugged in... with no system attached.

    The Kill-a-watt is a reasonable tool for simple testing (along with consistent comparisons) -- different memory, fans, drives and bios settings will have visible affects, as does the type of use.

    The C1E and EUP 2013 settings will make a few watts difference (if you're looking for cumulative effects). The ASRock H77 mobo may also be set by default for higher performance rather than energy use.

    I assume stefanqn's ASRock does not have a graphics card (given it was not in the profile) For low power, the integrated graphics chips are important. First, they are usually worth a minimum of 15-20W savings (given most graphics cards at normal use or idle) which is obviously significant. Second, if you use a small output PSU you really don't want a device like a graphics card that can spike at higher power. A 55W chip shouldnt go above its TDP :-) The SSD's will also knock a watt or 2 but modern platters and low rpm drives are pretty good. For server use, onchip graphics are convenient.

    I've also done undervolting (especially with the K10 AMD platform) which can make a nice difference on older machines (way too much headroom in the old quads... but then chip yields are based on statistics so you have the confidence interval in your favor). One of my more interesting surprises is how efficient some of the Intel dual and quad core systems could be (Intel seems more efficient by default, AMD energy consumption could be tweaked much more). I've got an 8MB Q9450 Gigabyte EP45 that idles at 67W including the HD5450 graphics.

    But with the new system/chips energy optimization is already built in... you really have to work at undoing it (as most of the overclockers will tell you). The best opportunity will be in matching output with efficiency. I've looked at a number of the 150W or less PSU's and very few 80+ models. And as we are discussing, many modern full desktop systems (or servers) can operate in the under 50W space...

    The trick is in finding good quality adapters that convert bricks into standard plugs for motherboards and cpus (and sata).
    Last edited by AcmeGadgets; 02-22-2013 at 04:05 AM.

  10. #10
    Join Date
    Dec 2009
    Location
    Austin TX
    Posts
    10

    Default Re: ASRock H77 Pro4-M - unreasonable high power consumption?

    Quote Originally Posted by profJim View Post
    I need to add a 3rd question:
    ...
    3. What is the maximum wattage that your system might need?
    Whatever the answer is to question 3, you'll need to add a safety fudge factor of 30 - 40% for your powerbrick or psu.
    Understood. As mentioned I don't use a graphics card for that reason -- too much variability in power demand. (and the built in HD graphics are actually pretty good for movies and such). With an integrated CPU I don't go above the TDP and the cumulative demands of the memory and HDD all stay within narrow ranges so you really do not need such a large safety factor.

    My systems don't really have any combined components that could go over 65 watts in practice (assuming everything was run at the max simultaneously...)

Page 1 of 4 123 ... LastLast

Thread Information

Users Browsing this Thread

There are currently 1 users browsing this thread. (0 members and 1 guests)

Tags for this Thread

Posting Permissions

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