digg

Join DiggAbout

Donkeys and Elephants and Delegates,oh my!
Check out the most popular Elections 2008 content on Digg.

See the original image at gizmodo.com

Solid State Drives Don't Extend Battery Life - Shorten's It

gizmodo.com — Solid state drives (SSDs) are the inevitable future of mobile computing, but a new experiment by Tom's Hardware is extremely disappointing. It ends up that the touted power savings of SSDs over their moving-parts-laden cousins are nonexistent. In fact, SSDs are sucking more power than conventional hard drives. How is this possible?

Bury
show profanity settings
expand all
  • tvanwyk, on 07/02/2008, -9/+195http://www.wikihow.com/Use-Apostrophes
    Just wondering what belongs to "Shorten."
    • kinglenster, on 07/02/2008, -3/+27I think it was 'it'
    • Hockey13, on 07/02/2008, -1/+17Maybe he was just referring to Hubert von Shorten's IT department. Have you ever thought of that?
      • OmarJasso, on 07/02/2008, -0/+11In that case, this kid needs to drop the apostrophe placement class and focus on capitalizing procedures.
    • VaporBro, on 07/02/2008, -6/+1Shrinky Dink; remember?
    • rodeosmurf, on 07/02/2008, -2/+15There's no need for an s anyway because the word shorten is referring to the plural, drives.
      • giveer, on 07/02/2008, -5/+1Actually, 'shorten' is referring to battery life, a singular trait of "drives", it's not referring it "drives" itself.
        If it was, the article would be about shortened hard drives - something that you could do with an HD and a sharp axe.
      • rodeosmurf, on 07/02/2008, -0/+2Shorten is a verb referring to the action of the drives. The drives shorten battery life, they don't shortens it.
    • graemee, on 07/02/2008, -6/+3e-penis. This article shortens SSD users e-penises.
    • zombies187, on 07/02/2008, -7/+1Petty.
    • galaxie, on 07/02/2008, -7/+1Wow, you really are an a**hole pal
    • SmokedL, on 07/02/2008, -0/+3From the real Toms article:
      "we took four SSDs that we had available in our test lab, and ran a series of Mobilemark benchmark runs on a Dell Latitude D630 notebook. We found runtime differences of up to one hour (!) when using a flash SSD compared to a high-performance 7,200 RPM 2.5” notebook hard drive."

      The irony of the first two sentences in that quote is just too much.

      As usual the crew at Toms are incompetent idiots. They run a benchmark continuously until power runs out. If you do this comparing an SSD to an ordinary drive the laptop with the ordinary drive will live longer. Possibly partially because the SSD draws more power under use, but mostly because since the SSD is far faster the CPU will run, rather than wait for the drive, a far greater percentage of the time consuming far more power. And that's not even beginning to take into account how utterly useless a comparison of drive power consumption under 100% load is to 99.99% of ordinary users.

      If you instead compared how many times the benchmark managed to run before power ran out that would be a far better metric and the SSD would win. And far better still would be to compare SSDs to ordinary drives under ordinary usage. Oh wait, that has been done and SSDs DO USE LESS POWER IN REAL WORLD USAGE.

      Buried as the usual inaccurate "Tom's paid for disinformation about Hardware". I guess they are taking money from some drive manufacturer this time around rather than from Intel.
      • kanabiis, on 07/03/2008, -0/+1Link to said proof? I mean independent verification proof, not Apples (or any other SSD manufacturer) in house testing.

  • tehhowch, on 07/02/2008, -3/+91Here's the direct link to the story on Tom's Hardware:

    http://www.tomshardware.com/reviews/ssd-hdd-batter ...
    • cnot3, on 07/02/2008, -5/+6I like how they picked the one graph out of 9 or 10 from Tom's Hardware which put the Hard Drive on top. In performance tests it nearly always lagged behind the SSDs.
      • tidu, on 07/02/2008, -2/+9That's the point of the submission... that it has better performance but they "Don't Extend Battery Life - Shorten's It" (sic) ;)
    • PATSCRU, on 07/02/2008, -0/+4smash the blogspam.
  • buba1243, on 07/02/2008, -3/+33Give SSDs time they will start to work on power saving technologies and then far surpass magnetic drives.
    • zadadka, on 07/02/2008, -1/+5They'll need to keep temperatures down.....that may be a struggle......
    • perspicuity777, on 07/02/2008, -3/+1Agreed, it couldn't possibly take much to fix this.
    • e2superman, on 07/02/2008, -7/+1And are you saying this because you know why they will eventually consume less power or are you just saying this because you are a duche looking to get dugg? Btw I agree however I hate when people just blindly support a technology.
      • Jerky1312, on 07/02/2008, -1/+2If you agree with him, what the hell was the point of your reply? Are you saying this because you are a deuche looking to get dugg down?
      • buba1243, on 07/02/2008, -0/+1No I am saying that because I work in the computer programming industry and try to keep up on the latest hardware to be able to do my job better. I also like the idea of using less energy and in my spare time read on battery usage and would like to go on to do some work in that field. Why not try to bring something intelligent to the discussion instead of just calling someone a "duche" (douche) cause they think that SSD are the way of the future.
      • e2superman, on 07/03/2008, -0/+1Interesting. You are a CS. Lame. If you were an EE I would consider listening. CS have no clue on electronics beyond fan boy reviews ;).
  • fluidfoundation, on 07/02/2008, -5/+43She's gone from suck to blow!
  • leerayIG88, on 07/02/2008, -6/+9Can you raid array SSD? Would be interesting to try...
    • dhughes, on 07/02/2008, -0/+12 Yes, it's just a drive, and I think Tom's did it, I've read about someone doing that but forget who it was if it wasn't Tom's. Check Ars Technica, may have been them.
    • exscape, on 07/02/2008, -2/+22Hell yes. :-)
      http://www.nextlevelhardware.com/storage/battleshi ...
      9 SSDs in RAID 0. Data safety? Pff. 820MB/s sustained read!
      • Foot56, on 07/02/2008, -2/+19 Hd's in raid 0 never heard of this. I thought they had to be even to stripe them?
      • Drakoola, on 07/02/2008, -2/+2I thought so as well, but never really understood why. It seems like it's just splitting the data into parts, but then again I don't know too much about file systems.
    • santaliqueur, on 07/02/2008, -4/+4It wouldn't be interesting, because it's basically just like any other drive.
    • verkon, on 07/02/2008, -1/+15SSD's fail on the "inexpensive" part.
      • Dubbsacc, on 07/02/2008, -2/+6So coin a new term "RAED".

        Why not, we need more acronyms.
      • dobesov, on 07/02/2008, -0/+7Actually, much like how THEY changed the V in DVD to "Versatile" the I in RAID is now considered "Independent"
    • rblancarte, on 07/02/2008, -1/+9Woa, can you imagine a Beowulf cluster of these?

      Oops, my bad, I thought this was Slashdot.
  • mweflen, on 07/02/2008, -4/+80Gizmodo blogspam shortens my personal battery life, too.
  • SebsWorld, on 07/02/2008, -7/+5Totally unexpected result. Future of hardware suddenly doesnt look as good as it did.
  • superkendall, on 07/02/2008, -12/+7I've seen other tests that measure Macbook Air battery life with SSD used, and all of them show somewhat longer life (though not usually a huge amount larger). Something may be wrong with his testing, or the system he tested the SSD drives with...
    • Onyxblaze, on 07/02/2008, -3/+5Two words: Macbook Air
      • cnot3, on 07/02/2008, -2/+9Yeah, that glorified playskool computer should never be used in a reputable benchmark.
      • superkendall, on 07/02/2008, -6/+3But if the Air is actually getting longer battery life for users with the SSD, then how is it not helping the user get extra battery life - the fundamental point Toms was trying to make. If there are cases where Toms assertion is false then something is wrong with his testing.
  • FredFredrickson, on 07/02/2008, -8/+3If solid state drives continue to deliver really bad results, with limited re-writes, huge cost, and now shorter battery life, why are they still considered the inevitable future of mobile computing?

    I mean, something better can come along... until then, traditional drives seem to be working just fine.
    • shaka999, on 07/02/2008, -0/+4If you look at the mean time before failure the SSD drives are superior to traditional drives. The re-write issue just isn't relevant.
      • FredFredrickson, on 07/02/2008, -1/+1Well what about the other issues then?
    • Hello1024, on 07/03/2008, -0/+1Has anyone reputable source actually done any reliability tests on these things?

      I hate to say it, but with good wear leveling and bad block replacement algorithms, I can't see how flash drives can't last 100's of years at least.

      The bad reputation flash drives got were from a few initial USB flash memory sticks with no wear leveling all all that all failed within a month or two - the industry learned it's lesson and doesn't do that anymore.
  • MarkDenali, on 07/02/2008, -0/+27Poster should just link to the source.
  • VIPAccess, on 07/02/2008, -6/+23I'm pretty sure this has been debunked. And who the ***** believes Tom's Hardware? Their creditability died back in '02.
    • burnedtubes, on 07/02/2008, -0/+11Agreed. Their testing methods leave a lot to be desired.
    • DeFex, on 07/02/2008, -0/+18what happened in 02?
      • sockpuppets, on 07/02/2008, -2/+20An undersea, unexplained mass sponge migration.
      • Malnilion, on 07/02/2008, -1/+4sockpuppets, the sponges migrated about a foot-and-a-half.
      • VIPAccess, on 07/02/2008, -0/+14There was some of manipulation of numbers between AMD and Intel in one of their articles in Intel's favor that proved to be false. At the time the site was plastered with Intel ads. It may have been early '03 I can't remember. But I'm pretty sure before Athlon 64.
    • stealthc, on 07/02/2008, -2/+7I'm with DeFex. What exactly happened in '02? I bought a laptop hard drive based on their recommendation and couldn't be happier.
    • ArgonTime, on 07/02/2008, -7/+1What happened on '02? Oh right, nothing. You're just making up things.
  • dhughes, on 07/02/2008, -3/+4 Even with the difference, if it's true and it depends on the situation.

    As for the an SSD always being 'on' no matter what compared to a regular spinning disk drive with platters I could see someone writing a script to unmount the SSD drive so it's off with your apps running in system RAM and then having the SSD drive remount when needed sort of like a hybrid car that stops its engine at a stop light. That would improve its performance power-wise although it would be horrible for the swap file on it.

    I would prefer an SSD anyway due to its solid state nature, I don't like rotating media, the same goes for magnetic media versus optical, an optical solid-state drive would be fantastic.
    • B1663r, on 07/02/2008, -1/+1So whats wrong with "rotating media"???
      • zombies187, on 07/02/2008, -1/+9Its a problem when it stops rotating.
    • caseycoold, on 07/02/2008, -1/+6moving parts+gravity=$#@!$!!
    • Hello1024, on 07/03/2008, -0/+1It would be hard to do the "unmount" at the OS level, but at the firmware/hardware level it should be relatively easy to implement. I guess now that this issue has been shown in public, the next firmware update for most types of drive will cut power consumption dramaticly. Within 2 years I guess flash drive power consumption for the average user will be halved.

      The fact is that most of the time the drive is idle, and with a flash drive, idle means it can be entirely powered off except the sata interface to listen for commands - total power consumption should be in tens of milliwatts. Power consumption when active may be harder to optimize, but remember since it should be able to switch between the states within milliseconds, it can go into "idle" mode pretty much immediately after every read/write operation.
  • cnot3, on 07/02/2008, -1/+12Large capacity solid state drives are still relatively new technology, I'm sure hard drives were less efficient than tape drives when they came out, too. Plus there is still the advantage of no moving parts. And that blogspam neglected to show that the access time on the solid state drives was less than 1% of the access time on the hard drive. Faster video cards eat up battery life, too, but no one would argue that we should all use Intel GMA.
  • Duositex, on 07/02/2008, -4/+35buried for gizmodo. Diggers are lazy. Finding only main stream sources for articles while countless sources exist. Do some footwork. Investigate. Stop submitting regurgitated article summaries as originals.

    Who am I kidding..
  • DeFex, on 07/02/2008, -2/+9how hard can it be to go to the bottom and find the original article.
    • mweflen, on 07/02/2008, -0/+3Gizomodo rakes in ad dollars for each view they get on their page, while Tom's Hardware gets squat if the person doesn't click through.

      Spam is spam is spam, any way you slice it.
  • zadadka, on 07/02/2008, -11/+9Try pronouncing "SSD" out loud, without sounding like a stammering Russian....
  • grexeo, on 07/02/2008, -2/+13First they claimed superior performance, but it turned out that random-access was substantially slower than conventional hard drives.

    Secondly they claimed SSD to be lighter, but on a per-GB comparison with modern capacities they're heavier, especially when compared to 500GB+ conventional drives.

    Thirdly they claimed SSD would be more reliable, but the current failure rate is substantially higher than conventional drives. Some manufacturers are seeing failure rates as high as 1 in 3.

    And now it turns out they use more power than conventional drives too.

    I don't doubt that SSD's will eventually replace conventional drives, but we're a lot further away from it then the manufacturers want you to believe.
  • Iwantawii, on 07/02/2008, -1/+22What about the fact that the CPU is busier because it's not being blocked by IO? The article doesn't mention this. They just stated that they "ran a series of Mobilemark benchmark runs on a Dell Latitude D630 notebook" without saying how many times each drive completed each test, or if they even counted. Maybe they just let it spin and went to lunch.

    But this is a huge factor, the article is mostly worthless without that info.
    • Hello1024, on 07/03/2008, -0/+2But it's approximately fair across different drive types, and won't make much difference since with DMA CPU load is minimal during IO, especially if it's directly to the disk and doesn't have to go through all the layers of OS translation (filesystem etc.)

      Having said that, toms hardware should've used a power meter on the disk power lines, to confirm where the power is being used.
  • seandfeeney, on 07/02/2008, -0/+5With solid state, I am not too worried about the battery life issue, I'd just plug the laptop in anyways. I want SSD's because 5200 RPM's suck!
  • sandiegodude, on 07/02/2008, -1/+17Take an external drive and 10 inch drop test it while its running.

    Now do the same with the SSD.

    SSD's have the benefit of no moving parts which makes them suitable for many applications where disk drives are too fragile.

    Power savings will come with time, you're talking brand new technology versus something that has been around for 40+ years.
  • john2kx, on 07/02/2008, -13/+7Anyone else read "Solid Snake" at first?

    State?
    Staaaaaaaaaaaaaaaaaaaaaaate!!!!
  • makenshin, on 07/02/2008, -1/+10I never thought SSD saved power. My interest in them is all for how fast SS is. OS and hardware designed around SSD usage could probably lower power consumption.
  • superkendall, on 07/02/2008, -3/+8"Shorten Is It"?

    Come on.
  • jabberwolf, on 07/02/2008, -8/+1I so Love Tom's website !

    He actually tests things out and makes comparisons. Kinda strange most places dont do that, especially when it comes to apple.
    Heaven forbid they actually make a head to head comparison using the EXACT same hardware (bus speed n' all).
    • chrispr0, on 07/02/2008, -0/+2Heaven forbid Tom's to use any sort of actual scientific methodology in the testing/reporting process.
  • Niz1, on 07/02/2008, -0/+4For me its not really about the power saving its about it not breaking when dropped as easily (ive dropped a 250GB bye bye from 2 feet) and also about speed . At the moment the cost the price for these data "luxuries" are just too great and also the capacity to price. 1TB HDDs cost about £100 if not cheaper the rest of the world, i cant imagine beating that price for storage, or even go that high?, maybe when im 40? (in 20 years)
  • rblancarte, on 07/02/2008, -0/+12I can't fully agree with the findings of the article. The main thing they did was run Mobilemark 07 until the laptop died. HOWEVER consider this:

    SSD is faster than magnetic hard drives, thus it can pass more data to the system, thus keeping the whole system under load more. It is possible that this data would keep the CPU loaded etc.

    Another thing, let's say that the test is getting data, running calculations, and saving data. The SSDs are faster, thus maybe over time they actually pass around twice the data of a magetic HD. Sure, Battery life is down 1 hour, BUT if you passed more data, then that 1 hour isn't lost in the same way.

    There is a flat out chart that shows:
    SanDisk drive power usage - idle - .5 watts loaded 1.0 watts vs a Hitachi magnetic drive 1.1 watts idle, 3.2 watts loaded
    Because of this, the results make no sense.

    I am not saying that their story isn't true, because I don't know.

    HOWEVER, I think it would be better to compare very like results to like results. IMHO, the above situations are possible. But how about a simple test - read data off the drive constantly and see which dies first.
    • kingmanic, on 07/02/2008, -0/+1I somehow doubt most laptop users do heavy number crunching on their laptop. IT's more for meetings, and taking some work away from the office. I'd imagine word, excel, power point, and explorer being the most often used function so increased performance isn't going to be a massive plus for laptops. SSD certainly reduces seek times reducing the hit to the system but I'm not sure if it would double performance as throughput is about the same.
      • rblancarte, on 07/02/2008, -0/+1To be fair, I am pulling these numbers out of my butt. I don't know one way or the other the throughput of the drives.

        But consider this test:
        Program that does nothing but read and write data in different ways from drive (sequentally and random access), CONSTANTLY, and then reports/records throughput. Then when the battery dies we know:
        1. most of the battery consumption was due to the drive performance
        2. how much data is passed, thus we can try to get a bytes transfered per watt figure

        ultimate, if bytes/watt is significantly higher for one format or the other, then you would have a much better indicator of battery life of each drive type.
      • kingmanic, on 07/02/2008, -0/+1SSD tends to seek faster so small non adjacent files are less of a problem. Throughput is limited by the device bus for both HD and SDD. SATA peaks at 3gbits/s (~375MB/s) while Samsunds latest SSD has awrite speed of 200 MB/s. Assuming SATA's average is a bit lower then SSD is comparable to SATA 3. The only real difference is seek time which is more to do with performance on small non localized files.

        Writes are also asymmetric on SSD. read fast but write slow. HD's are the same both ways.

        The only real trump is seek time which may make SSD a great middle cache between HDD and Ram.
  • harry8227, on 07/02/2008, -1/+0Relax, its not
  • RusskiGuy, on 07/02/2008, -0/+4I don't like the experiment, because it apparently compared various SSD's to only one conventional HD. If power consumption varies for SSD's, I would assume it also does for HD's. I wouldn't be surprised if some HD's in fact consumed more power than some of the more efficient SSD's.
  • tnoy, on 07/02/2008, -2/+3Mondern laptop drives only use 500mA or less, they're not the source of poor battery life.
    • Ramble, on 07/02/2008, -0/+1Ampage is irrelevant unless in context with voltage.
      • tnoy, on 07/02/2008, -0/+4Yes, and all laptop drives have been using the same voltage for over a decade.

        Sorry for not doing the math for you.
    • e2superman, on 07/02/2008, -0/+2Agreed. It is the backlight. When LED backlights become standard then the next thing is likely the CPU (if not already) is Watts used.
  • AllyOfReason, on 07/02/2008, -0/+6Buried as inaccurate.
  • Baryn, on 07/02/2008, -3/+2FUUUUUUUUUUUUUUUUUCK

    I just spent $2800 on a high capacity SSD! :knocks over magazine rack:
  • thespanielator, on 07/02/2008, -0/+44Slasdot is running the same story, the following is lifted from one of the comments:

    "Unfortunately, the article comes to miserably faulty conclusions, and nobody seems to have noticed it. One thing everybody needs to note is this: the test that Tom's ran was designed to run a program to completion over and over again. All that Tom's came up with, thanks to that test, is that the computer with the SSD ran out of power faster. What they SHOULD have noticed, and what any sane and logical tester would have done, is counted the number of times the program RAN. A system with an SSD will run the program a significantly higher number of times because of its increased speed and lack of moving parts. With the higher number of runs, the CPU STAYS MORE ACTIVE, sucking more power out of the system and thus causing it to power down earlier. The only reason the HDD system died later is because the CPU idles waiting for it to retrieve data. The SSD may have chewed up your battery faster, but it did, comparatively, run through that program a BUNCH more times. I am willing to bet the work per watt was much better out of the SSD system than the HDD - and once again, Tom's Hardware's scientific testing system proves itself anything but."
  • chrispr0, on 07/02/2008, -0/+7The testing methods Tom's used do not prove the hypothesis & headline. If you need this explained, read the comments on Tom's. Buried for ignorance.
  • drunknmunky1, on 07/02/2008, -0/+4Could Tom’s Hardware be Wrong?

    No, our results are definitely correct.

    ... =/
  • DifferentAngle, on 07/02/2008, -0/+4Tom's Hardware is a garbage site. Every one of their "experiments" are mythbusters quality.
  • kettlehead, on 07/02/2008, -0/+1Holy *****, here comes an S!
  • bincoder, on 07/02/2008, -0/+1I dont own a laptop but in my less than one year old camcorder which uses 2 AA batteries (NiMH rechargables) and currently a 2 gig SD card for storage, the batteries will get good and toasty indicating vast reserves of energy being used.
    A recorded file of around 1 gigabyte takes almost a half hour to transfer to the pcs hard drive. The PC itself can transfer a similar file from one location to another at least ten times faster using the quaint olde fashioned hard drive.
    So still I wonder why do these cards have a reputation of being blazingly fast and power efficient?
    No failures yet though on the Kingston SD card with almost daily use.
    • deadbaby, on 07/02/2008, -0/+1You're probably using a slow SD card and you're using it over USB which is far less efficient for mass storage than a dedicated SATA controller.
    • Hello1024, on 07/03/2008, -0/+1The SD card interface is pretty slow - it only does 4 bits per clock, and runs at 20Mhz, so even if everything else was ideal you'd still only get 10Mbytes per sec out of your card.

      SD is an old standard now anyway - everything in the future really ought to be using the USB interface directly.
  • jcswright, on 07/02/2008, -0/+2This doesn't tie in with http://www.samsung.com/global/business/semiconduct ...
    Something doesn't add up.
  • AmericansRevolt, on 07/02/2008, -0/+1am i the only one who doesnt care? harddrives never were a huge concern on my power bill as it is
  • OmegaWolf, on 07/02/2008, -0/+1It makes sense logically that the battery life would be shorter with no variable in power consumption.

    What about that racetrack storage I read about back in April? Isn't that superior to both magnetic drives and SSD? Won't that be the inevitable future for laptops?
  • pHr34kY, on 07/03/2008, -0/+1I thought they would have hooked a multimeter up to the HDD power cable and measured the current going through it. Measuring the entire system while only changing one component seems a little error-prone.

    In fact, I'm sure Tom's have done this test before, and I did it exactly how I described above, and the SSD was a LOT more efficient.
  • MrRobotoSki, on 07/03/2008, -0/+1Tom's is slanted and tested the PC's that were using SSD's had more components and higher power requirements. Check out their advertisers. You can be sure that Western Digital is one of them. But Western Digital doesn't have a personal stake in any of this.

    Seriously though, Tom's Hardware is not the great hardware review site it once was, it's owned by "Best Of Media". Their is no more journalistic integrity left. They test $8,000 PC's using games like Serious Sam and Prey.

    Whoa you mean Serious Sam gets 270 FPS on that top of the line Falcon Northwest! Truly Amazing!
  • Atomic1fire, on 07/03/2008, -0/+1In order to make a decent test you need to test a wide range of hardware
    A serious gamer (high end stuff)
    A casual web machine (only has as much as needed to view web/check email and maybe freecell)
    an Office PC (fits as far as office apps and other related sofware and requirements)
    and you need to accommodate a wide range of hardware to do that.
    • nephilimx, on 07/03/2008, -0/+1Hardware isnt that important, but they do state if HDD isnt defragged, it would use more power due to data beinf scattered
  • kanabiis, on 07/03/2008, -0/+1Well the jury seems to be still out on SSD failure rates. It was widely reported that SSD drives were having quite a high, some say 20-30% failure rate, Dell claims the study was full of *****. Which of course would be expected, since Dell is in the business of selling SSD drives.... I think it is way to early to declare victory over traditional hard disks.

    SATA drive manufacturers claimed victory over SCSI disks way too prematurely, and we know how that turned out, SCSI disks are still the most reliable drives available.
    • Hello1024, on 07/03/2008, -0/+1what typeof interface it has isn't a significant causation factor in reliability. It's simply because in general businesses use SCSI for performance and because all their old systems use SCSI, and businesses also require reliability. Therefore drive manufacturers pt more effort into reliability when making scsi drives, and couple that with the higher price tag as well.
  •  

    Login or register to add a comment

    Create a new account or login to join in the conversation on Digg. You'll also be able to Digg stories to help promote things you like.


© Digg Inc. 2008 — Content posted by Digg users is dedicated to the public domain.
DIGG, DIGG IT, DUGG, DIGG THIS, Digg graphics, logos, designs, page headers, button icons, scripts, and other service names are the trademarks of Digg Inc.