logo elektroda
logo elektroda
X
logo elektroda

HDD vs SSD - we measured the power consumption.

TechEkspert 108327 41

TL;DR

  • Measured power consumption of HDDs and SSDs on SATA power rails, comparing 3.5", 2.5", NAS, laptop, and SSD models for home NAS and portable builds.
  • Current was measured on 5V and 12V branches with multimeters, and waveform shape was checked on an oscilloscope across a 0.1Ω resistor.
  • WD RE4 2TB drew 8.8W idle, 10.8W during sequential write, and 11.4W while seeking, with a ~1.8A start current.
  • 3.5" NAS HDDs usually hovered around 10W, 2.5" laptop drives stayed below that and often needed no 12V rail, while SSDs idled as low as 0.1W.
  • Maximum-speed tests may overstate typical use, because quieter operation can consume less power, and multi-drive arrays need extra startup current planning.
Generated by the language model.
ADVERTISEMENT
Treść została przetłumaczona polish » english Zobacz oryginalną wersję tematu
📢 Listen (AI):
  • #31 14577691
    TechEkspert
    Editor
    Posts: 7031
    Help: 16
    Rate: 5452
    Board Language: polish
    It also caught my attention, perhaps apart from the logic of the disk, the energy is consumed by the HDD head positioning coil, which has to perform positioning, regardless of whether there is a read or write occurrence, it may confirm an increase in energy consumption during searching, i.e. an extreme case of intensive operation of the for some disks, eg WD RE4 2TB, the search resulted in a significant increase in current consumption in the 12V branch - positioning coil?, and intensive sequential write / read in the 5V branch - logic?).
  • ADVERTISEMENT
  • #32 14577915
    freebsd
    Level 42  
    Posts: 6790
    Help: 766
    Rate: 2025
    Board Language: polish
    TechEkspert wrote:
    which must perform positioning, regardless of whether there is writing or reading, this can confirm an increase in energy consumption during searching, i.e. an extreme case of intense work of the positioning mechanism

    I have similar thoughts.

    TechEkspert wrote:
    interestingly, for some drives, e.g. WD RE4 2TB, the search resulted in a significant increase in current consumption in the 12V branch - positioning coil

    Higher voltage -> greater force? (and lower currents to control by the way?) This disk is supposed to be fast, maybe it's one of its secrets? :-)
  • ADVERTISEMENT
  • #33 17450823
    amadeusrecords
    Level 2  
    Posts: 4
    Board Language: polish
    Respect!
    I am interested in the short moment of power consumption when switching on the disk. This is probably a very increased consumption for a very short period of time.
    Well, I deal with sound recording and use the RME UFX interface. It has the additional option of connecting a USB flash drive as a parallel recording system. Unfortunately, the very slow transfer puts into question the recordings with the use of very many traces, there are transmission errors.
    Contrary to my destiny, I used a 2.5 HDD for this, until after 4 years the interface power supply crashed when turned on.
    Now I'm thinking about using an SSD there, but I'm still afraid of repeating the situation.

    Best wishes,
    Tomasz Lida
  • #34 17451466
    TechEkspert
    Editor
    Posts: 7031
    Help: 16
    Rate: 5452
    Board Language: polish
    When starting, hard drives can consume more electricity when, for example, they unscrew the plates,
    large 3.5 "drives with multiple platters could even be configured to slowly unscrew platters,
    it mattered when, for example, the matrix started up with 24 hard drives.

    In smart there was even an error counter called "buzzing" and it concerned a problem with starting the platter drive, it could be caused by problems with the power supply, with the drive itself.

    For mechanical drives, you can think of an external power supply, e.g. from a power bank.

    In the case of SSDs, I did not notice such a problem, interestingly there are miniature SSDs with probably a native USB interface, unfortunately they are quite expensive https://www.komputronik.pl/product/546363/adata-sd700-256gb-ssd-zolty .html

    You can also look for a few manufacturers of flash drives with SLC memories, they should provide good write speeds:
    https://www.komputronik.pl/search/category/1?query=pendrive%20SLC
  • #35 17451939
    amadeusrecords
    Level 2  
    Posts: 4
    Board Language: polish
    Thank you very much and regards! :-)
  • ADVERTISEMENT
  • #36 17453139
    TechEkspert
    Editor
    Posts: 7031
    Help: 16
    Rate: 5452
    Board Language: polish
    If you manage to conduct experiments with recording multiple tracks, it is worth appearing in the articles https://www.elektroda.pl/rtvforum/forum41.html
    the topic is not obvious and it is worth presenting your achievements.
  • #37 17454012
    amadeusrecords
    Level 2  
    Posts: 4
    Board Language: polish
    Thank you very much for inviting me :-)
    Regards!
    Tomasz Lida
  • #38 18138846
    True_Brujah
    Level 13  
    Posts: 26
    Help: 3
    Rate: 6
    Board Language: polish
    As you can see, it all depends on the hard drive and its production date. I have such an invention of the Samsung MZ-5PA064A SSD. It clearly says 5V 1.6A. I wonder how it would come out with the measurements.
  • #39 18139815
    TechEkspert
    Editor
    Posts: 7031
    Help: 16
    Rate: 5452
    Board Language: polish
    @True_Brujah high current, are you able to test it in practice?
  • #40 18149298
    True_Brujah
    Level 13  
    Posts: 26
    Help: 3
    Rate: 6
    Board Language: polish
    At the moment I have measured the current on 2 disks and it does not look so rosy.

    Test computer with SATA II connector (unfortunately I had nothing else at hand)

    SSD Samsung 470 64GB (SATA 3.0Gbps)
    Sequential Write: 0.488A
    Sequential read: 0.304A

    SSD PNY CS900 120GB SATA III (6Gbps)
    Sequential Write: 0.160A
    Sequential read: 0.149A

    There is information from the data on the disks:
    PNY 5V 1A
    Samsung 5V 1.6A

    ------------------------------------- Little Update ----------- --------------------------
    SATA III test computer

    SSD Samsung 470 64GB (SATA 3.0Gbps)
    Idle: 0.050A
    Sequential Write: 0.505A
    Sequential read: 0.343A

    Samsung 850 EVO
    Idle: 0.167A (system disk, that's probably why the consumption is so high)
    Sequential Write: 0.361A
    Sequential Read: 0.396A ??? (test performed 3x)

    PNY
    Idle: 0.057A
    Sequential Write: 0.198A
    Sequential read: 0.181A
  • #41 18151478
    TechEkspert
    Editor
    Posts: 7031
    Help: 16
    Rate: 5452
    Board Language: polish
    What transfers have these SSDs achieved with these power consumption?
  • ADVERTISEMENT
  • #42 18186574
    True_Brujah
    Level 13  
    Posts: 26
    Help: 3
    Rate: 6
    Board Language: polish
    SSD Samsung 470 64GB (SATA 3.0Gbps)
    Sequential Write: 0.488A - 168MB / s
    Sequential read: 0.304A - 252MB / s
    This is an old SATA II drive, so the results should come as no surprise.

    SSD PNY CS900 120GB SATA III (6Gbps) - purchased 09.2019
    Sequential Write: 0.160A - 260MB / s
    Sequential read: 0.149A - 370MB / s
    As you can see, a cheap drive for PLN 79 is fine, but it lacks an expensive brand drive.

    Samsung 850 EVO SATA III - Purchased at the end of 2015
    Sequential Write: 0.361A - 491MB / s
    Sequential read: 0.396A - 507MB / s
📢 Listen (AI):

Topic summary

✨ The discussion centers around the power consumption measurements of HDDs and SSDs, particularly in the context of building devices with SATA connectors, such as NAS systems. Users share insights on the energy efficiency of SSDs compared to HDDs, noting that while SSDs can consume similar power at maximum capacity, they often complete operations faster, potentially leading to lower overall energy use. Concerns are raised about the accuracy of the measurements due to the limitations of the multimeters used, with suggestions for using more precise equipment. The conversation also touches on the differences in power consumption during read and write operations, with some users noting unexpected results regarding energy usage patterns. Various brands and models of SSDs and HDDs are mentioned, along with discussions on measurement techniques and the implications for device design.
Generated by the language model.

FAQ

TL;DR: 3.5″ WD RED HDD idles at 3.4 W [Elektroda, TechEkspert, post #14528907] while a 128 GB SSD idles at 0.4 W [Elektroda, TechEkspert, post #14528907]; “SSDs finish the operation faster” [Elektroda, TechEkspert, post #14531343]

Why it matters: Lower idle draw and shorter tasks mean longer battery life and smaller power-supply bills.

Quick Facts

• 3.5″ NL-SATA HDD start-up surge: ~1.8 A @ 12 V [Elektroda, TechEkspert, post #14528907] • 2.5″ NAS HDDs use 0 A on 12 V rail; 5 V only [Elektroda, TechEkspert, post #14528907] • GoodRAM 240 GB SSD idle power: 0.1 W [Elektroda, TechEkspert, post #14528907] • SSD peak read: 394 MB/s at 2.4 W [Elektroda, TechEkspert, post #14528907] • Typical multimeter class UT58D accuracy: ±2 % + 50 mA [Elektroda, SStaszekk, post #14535549]

Which device uses less power at idle, HDD or SSD?

A modern 240 GB SSD idles at 0.1 W, while the most frugal 2.5″ HDD still needs 0.5 W [Elektroda, TechEkspert, post #14528907] SSD wins by a 5× margin.

Are SSDs always more efficient during heavy transfers?

Not always. At full-speed writes an SSD drew 3.1 W, similar to a 2.5″ HDD at 3.1 W [Elektroda, TechEkspert, post #14528907] SSDs finish sooner, so total energy per file remains lower.

How high is the inrush current when a 3.5″ HDD spins up?

NL-SATA drives hit ~1.8 A on the 12 V rail during spin-up [Elektroda, TechEkspert, post #14528907] A weak supply can brown-out USB or SBC systems—an edge-case many builders overlook.

Why can read power equal or exceed write power on HDDs?

Head positioning and platter drag dominate power. They act during both operations, so reads on some drives drew 0.58 A, matching writes [Elektroda, freebsd, post #14573871]

How do I measure my own drive’s power?

  1. Insert a calibrated ammeter in the 5 V and 12 V leads.
  2. Log current during idle, sequential read, and write.
  3. Multiply by measured rail voltages for watts. Keep leads short to reduce drop. [How-To synthesised from thread]

What factors raise HDD power aside from RPM?

More platters, faster head seeks, and 24/7 firmware lift current draws; NAS-rated 3.5″ drives used 4–11 W depending on mode [Elektroda, TechEkspert, post #14528907]

Can acoustic management lower HDD power?

Yes. Reducing seek aggressiveness shortens coil drive pulses, trimming 0.2–0.5 W according to datasheet figures [WD, SpecSheet].
Generated by the language model.
ADVERTISEMENT