11
May
EXCLUSIVE: AndroidPolice.com's Nexus One Is Running Android 2.2 Froyo. How Fast Is It Compared To 2.1? Oh, Only About 450% Faster
Last Updated: June 5th, 2012

Upcoming Frozen Goodness

Folks, we have a Nexus One. We also have Froyo. In fact, we have the two mixed together in the most delicious frozen-yogurty Android 2.2 kind of way you can imagine.

Not much has been officially released by Google about Android 2.2 - probably less than any other release to date. We all have a hunch about some new features but Google wouldn't officially confirm or deny any of them, other than the one about Flash 10.1.

They're saving it for May 19th-20th, and we're OK with that.

However, one of our team members has been playing with a Nexus One running Froyo for about a week now, getting more and more visibly excited every day.

Before we proceed, however, here's a quick picture to squelch any non-believers. These are not fakes and we will be playing with the phone on video very shortly.

EXCLUSIVE: AndroidPolice.com's Nexus One Is Running Android 2.2 Froyo. How Fast Is It Compared To 2.1? Oh, Only About 450% Faster

Yes, it runs Flash, and it runs it very well (we'll demo that separately but you can get the idea here).

As you've seen it already, it has a new home screen menu but we won't talk about that today.

Froyo 2.2 Performance

Today, we would like to point out the results of some speed tests we've been running on Froyo, using a tool called Linpack.

These results are so impressive, most likely due to the introduction of the long awaited JIT compilation into the kernel, that we felt they deserved immediate attention.

Linpack

Linpack for Android was specifically created to test the performance of Dalvik VM which, in short, is the heart and the brains of Android.

The benchmark calculates how fast your phone can operate, which can then be used to compare various phone models' relative performance. Linpack allows users to publish these results to its online tracker, but we'll talk about that a bit later.

Here's a brief introduction to Linpack by the authors themselves, with interesting parts bolded:

The LINPACK Benchmarks are a measure of a system’s floating point computing power. Introduced by Jack Dongarra, they measure how fast a computer solves a dense N by N system of linear equations Ax = b, which is a common task in engineering. The solution is obtained by Gaussian elimination with partial pivoting, with 2/3*N3 + 2*N2 floating point operations. The result is reported in Millions of FLoating-point Operations Per Second (MFLOP/s, sometimes simply called FLOPS).

This test is more a reflection of the state of the Android Dalvik Virtual Machine than of the floating point performance of the underlying processor. Software written for an Android device is written using Java code that the Dalvik VM interprets at run time. 

What is the purpose of this app?
This is a simple benchmark test to show performance relative to other phones for a standard calculation.  Linpack has been used for years on all types of computers, with a version used to rate the TOP500 computers in the world.

What speed is better?
A higher number is better.

Does having faster speed improve the android phones or what?
Yes, it should.  The Dalvik VM has a huge impact on the Linpack number.  A better number on the same device would indicate that a new version update has improved performance.  Or it could show that something has gone terribly wrong if the number goes down.

But enough about Linpack - let's take a look at how Froyo performs.

Linpack Scores

Just so you have an idea of the kind of scores you can expect from the benchmark, let me list a few sample numbers:

  • the Nexus One running Android 2.1 gets about 6.5-7 MFLOPS
  • my HTC Hero averages a measly score of about 2 MFLOPS (the phone is pretty slow indeed)

Here is a screenshot of the last test I ran on the Hero:

HTC Hero Linpack numbers

Froyo Linpack Score

So how does the Nexus One with Froyo do?

EXCLUSIVE: AndroidPolice.com's Nexus One Is Running Android 2.2 Froyo. How Fast Is It Compared To 2.1? Oh, Only About 450% Faster

A whopping 37.5 MFLOPS.

Think about this for a second. 37.6, compared to 6.5-7 - that's about a 450% speed improvement.

Conclusion

Things are starting to finally come together. Flash and huge performance gains, all in the same release. Flash is CPU hungry, so Android makes everything about the environment more efficient.

Do you see it? Adobe wasn't giving Apple what they needed, and Apple wasn't giving Adobe, well, anything.

But Google thought of the problem outside the box, like the ingenious engineers to the core that they are and made Android so much faster than it could finally run full Flash without a hitch.

Oh yeah, the phone is indeed noticeably faster.

Some of you (like Cyanogen and Uhhhh) pointed out that JIT compilation only helps programs that are not compiled to native code, which Flash and some video codecs/drivers seem to be. Thank you for this clarification.

While real-life improvements will most certainly not be 450% across the board, we still expect the N1 and even Flash to benefit from JIT because, as other programs become more efficient, more CPU will be available to such natively compiled programs.

These mysterious numbers from the Linpack Top 10 don't look so fake anymore now, do they?

EXCLUSIVE: AndroidPolice.com's Nexus One Is Running Android 2.2 Froyo. How Fast Is It Compared To 2.1? Oh, Only About 450% Faster

Artem Russakovskii
Artem is a die-hard Android fan, passionate tech blogger, obsessive-compulsive editor, bug hunting programmer, and the founder of Android Police.
Most of the time, you will find Artem either hacking away at code or thinking of the next 15 blog posts.

  • http://webofandroids.blogspot.com/ vkelman

    That's exciting, but how did you get 2.009 MFLOPs on your Nexus One 2.1? I've just downloaded Linpack, run it and got 4.761. I'm on AT&T network, but it shouldn't matter, right?

    I don't know how to get a screenshot on my Nexus, unfortunately.

    • http://www.AndroidPolice.com Artem Russakovskii

      Actually, I never said 2MFLOPS was on the N1 - it was on my HTC Hero. The N1 was 6.5-7MFLOPS.

      • http://webofandroids.blogspot.com/ vkelman

        OK, my mistake. Hero image and Nexus 2.2 image are located together, and there is no Nexus 2.1 image nearby - it was easy to make a mistake in 3:00 am :)

        But then, why my is Nexus One so slow? (4.76 vs. 6.5-7)

        • http://www.AndroidPolice.com Artem Russakovskii

          Yeah, sorry for the confusion - didn't have the N1 image handy, so had to go with Hero only.

          Not sure why your N1 is so slow, perhaps it's running a bunch of stuff in the background?

        • Brantyr

          Run the test again and again, takes a few runs to get everything cached for maximum speed. Putting it into aeroplane mode helps make sure the CPU isn't interrupted to respond to network traffic as well.

          And obviously make sure nothing else is running in the background.

      • http://webofandroids.blogspot.com/ Vladimir Kelman

        Artem, is there some kind of limitation on a depth of the thread here? I wanted to reply to your image below, but it has no "Reply" link to me.
        Anyway, I rebooted my N1 phone and measured MFLOPS again. At first it gave me even less (about 4), but the next moment I measured it again and got 6 MFLOPS. So, it wary greatly, not sure - why.
        And yes, I have 3 security apps running in the background (Lockout, Wave Secure, and Mobile Defence) mainly because I couldn't choose one and trying to evaluate them.

        • Salman Hasan

          My nexus one reads 6.8 MFLOPS. I Have not switched to airplane mode or anything. All I do is I use ES File Explorer and use the task manager to kill Applications; I have about 106 MB free on my RAM

      • Harsha S

        Did U try 2.2 on HTC Hero. If not why not..? Try porting it to HTC Hero or Leak that ROM to some Devs..?? Would be a nice feeling to here that it runs on Hero too with higher MFLOPs.. what say..?

    • Kr0nos

      The Screen is from the Hero.

      I hope 2.2 comes to my Magic // MT3G - by SuperF or cyanogen

      And I just flashed cyanogen 2.1 ROM yesterday :P

    • http://binary-lab.blogspot.com Vanquish

      The score 2.009 MFLOPs is from his HTC hero, not from N1.

  • Uhhhh

    So, fp operations are faster ... that's about all this says.

    Flash is very unlikely to be running within dalvik (it'd be running native under Linux) hence it'd be unaffected by these improvements...

    • http://www.AndroidPolice.com Artem Russakovskii

      Cyanogen pointed this out earlier today - as I'm not the biggest pro in low level Android implementation, I didn't know this - a correction/update will be posted ASAP.

      But such a boost in floating point operations is very important, don't you think? Think games and apps that will now be faster.

      • Gambit

        Except, of course, that most games aren't implemented in Java (Well, *I* am, but I'm provably nuts) and those that are, don't use FP, they use fixed point mathematics.

        Sorry.

  • http://www.nexusoneforum.net Nexus One Forum

    Wow awesome testing. Will link to this post with a front page article tomorrow! Thanks.

  • Sachin

    if this is true i will sell my iPhone on ebay

    • piprog

      I would probably first compare apples to apples. Android is a Java platform, while iPhone OS is native. This current improvement most probably only means that Android performance was up until now so terrible, that such an improvement was possible (although respect to the engineers achieving it!), and is now approaching iPhone performance (from the rear, I suppose).

      • JJJ

        What are you smoking piprog...??? I've got the latest iPod Touch and an HTC Incredible (do you even own an Android device?). I used to have a WinMo phone (Samsung Omnia) and I liked using the iPod Touch for email and browsing instead of my phone. However, now that I have the Incredible I use it for email and browsing since it's faster and does a better job than the iPod Touch...and that's with Android 2.1.

        • piprog

          hi JJJ, the Incredible is state of the art, no arguments about that. Will be interesting how iPhone 4 with matching cpu and with the new iPhone OS 4 will stack up (performance wise; no chance feature wise, though), but it is true that now there is at least one Android phone that is a real contender, which is good for both platforms, I'd say :-)

      • Jimlad

        lol piprog, apple os was designed for the hardware, considering that the iPhone has it's bollocks cut off to make it appear faster then it is a mighty thing that android is so much more capable and faster t'boot :p

        • piprog

          hi Jimlad, more capable: sure (v.s. non-geek usability;-). But faster? Is there any conclusive evidence out there? Of course Incredible & co now has a much faster processor, but on similar hardware e.g. Safari seems to be faster.

  • John Thompson

    Flash may not be directly dependant on FP operations, but it is a significant increase that will see the speed of many apps increase.

    Definitely great to see these behind the scene improvements coming with Froyo, bring on May 19th!

  • http://deftek.net tripplethrendo

    Will i be able to get 2.2 on the Eris?

    • dextroz

      go back to sleep. you are just getting 2.1. eris will never got 2.2 since it's end of life is next month.

  • cece

    Can you see if RAM usage also increased with JIT ??
    Might be the problem with first gen phones with low RAM available.

  • Jason Porter

    The way this test is done for MFLOPS is interesting for a "top of the stack" bench, but doesn't tell us much about the actual capability of the hardware.

    If you get a moment, can you see if it's possible to check the kernel log for the BogoMIPS value? On most linux systems this is at /var/log/messages but it may be somewhere odd on an Android system. BogoMIPS is evaluated for the processor(s) each time the kernel is loaded, and it's part of the default output log. It's not a particularly meaningful number in terms of application performance, but it will give a degree of comparability for the hardware itself.

  • Frank

    Downloaded linpack for my Nexus One and got MFLOPS 6.691

  • Richard

    Great! If Android Police has a build of Android Froyo, can they spoil the other features of Android Froyo before it releases?

  • Grashalm

    The Just in Time Compiler is awesome stuff. I always thought they have it already ;-)

    Desktop Java was known to be really slow, but then came the JIT and benchmarks were somewhere near C++.

    Not directly compiling to native code is great, since it enables performance optimizations even for already (bytecode) compiled applications. Also it can leverage all the target CPU options without recompling. (Mobile CPUs compared to desktop CPUs are pretty different)

  • Oscar

    Maybe the increased performance of floating point operations are due to android finally being able to access the NEON and VFPv3 instrucions on the Snapdragon (called VeNum)? If that's the case, flash, video decoding and everything else will run better as well.

  • Rogerio

    This is the iPHONE Killer!

    • zeagus

      Oh, hey, there's this bridge I have for sale...

  • phutwo

    These numbers are really low. Even if you don't use SIMD/Neon for floating point operation, you should be roughly 1 MFLOP per 1 MHertz. So for this phone it should say 1000 MFlop.

    • Jason Porter

      This isn't a measure of the hardware's performance. This is a Java-specific benchmark used to measure the actual (observable) performance of the entire software stack, including the Dalvik VM. Remember, this is LinpackJava customized for Android, not Linpack in the traditional sense. If they were running real Linpack (in C) compiled for the underlying linux kernel, they'd be far closer to the hardware and these numbers would be much higher.

      • http://eamon.nerbonnne.org Eamon Nerbonne

        ...and in any case it's simply not true that a 1MHertz processor can run linpack at 1mflop - even under the best of circumstances. Not all processors run 1 flop in 1 cycle (commonly, they don't) - some require extra latency between dependant read/writes, then there's the memory bandwidth and latency issues. It could also well be *faster* on a superscalar architecture with multiple ops in flight. Instruction mix can be a problem. In short, with a 1mhertz processor, you'd have an estimate for the order-of magnitude - but if you scored just 0.01Mflop on a 1MHertz machine, that might be odd, but not necessarily inconceivable.

        So yeah, using a 1GHz cpu and scoring 6-7 mflop in java on a phone running a GUI and network stack etc. that's not designed or optimized for linpack... ...is entirely reasonable, and not necessarily indicative of other performance metrics.

        • phutwo

          So I was quite interested in why the numbers are so low (JVM+JIT is close to C). In x86, a fully-pipelined FPU, you can get 90% of theoretical performance depending on cache speed/size (or a factor of 2 if you don't have fused multiply/add).

          However, I stand corrected. Cortex A8 has an unpipelined floating point coprocessor! The VFP takes 18-21 cycles for an FMAC. So for a 1GHz processor, 40MFlops. So wow, if they are using the VFP in the JIT, they're hitting theoretical numbers and the interpreter/Java is free (bottleneck is fpu)!

        • xvigour

          sorry but what you say is untrue, or I don't get your point.

          a core i7 (and cores before) is 4 flops/cycle.
          and a linpack can be 90+% efficient.

          so 1flops/cycle is easily reached.

          check the http://www.top500.org to be convinced

  • Brantyr

    37/7 ~= 520% not 450 (maybe leave the maths to the CPU :P)! Any way you slice it it's a crazy increase in speed, even running a JIT hack on cyanogenmod my Magic only goes from 2 to 3.6 MFLOPS - a rather paltry 180% increase in comparison!!!

    • http://www.AndroidPolice.com Artem Russakovskii

      Brantyr, 37/6.75 is about 550% but since we mention "faster", the original 100% is not counted, producing 450%.

      Think about what "100% faster" would mean and you'll see what I'm talking about.

      • Brantyr

        Whoops, I see what you mean now, sorry!

  • dj

    any noticeable changes to battery life under 2.2?

    have they reduced the amount of power needed by Android or is that just wishful thinking??

    • http://www.AndroidPolice.com Artem Russakovskii

      Unfortunately, we cannot discuss this at the moment. You guys will probably have to wait until Google publicly announces Froyo.

      • dj

        yeah, I thought that might be the case. thanks anyway

  • Adam

    It's worth pointing out that Flash is not exactly the CPU hog, it is rich content in general. Look at some of the HTML5 demos on the Nexus One and let me know how that goes. I'll give you a hint, choppy typically (same, worse really, for the apphole product). So yes Flash does use a large amount of CPU but it is more based on the content being pushed through the application and not the application itself. Now if they opened up hardware decode for Ogg and h.264 we'd see some wicked fast stuff (not to mention better battery)

    • Somebody

      Actually, flash IS just a turd.

      Try this: Measure CPU usage doing something with flash, and then measure CPU usage doing the *same thing* with native code. See the huge discrepancy?

      Flash is just a SLOW, UGLY memory hog, and has no place in this world.

      You can be assured that if flash shipped with AOSP or anyone's build of AOSP that I was interested in trying out, I would quite intentionally DELETE that component.

      • Adam

        Nice job giving concrete examples. Further on the phone nothing is Native.

        If by native you mean C or other compiled code without a VM then of course it will run faster. If I have an API to the hardware for decoding for example (something safari did but other did not until now on the Mac) I would run faster. The whole notion of a VM implies things will not run as fast, you don't get abstraction for free. This is the issue with Droid itself half the time. Is there room for improvement in flash, yes, is flash a pig compared to similar products for the web, find me a similar product and we can talk b/c right now HTML 5 in the majority of the browsers is a pig too.

  • Mike

    Will the Incredible with HTC sense be upgradable to froyo?

    • http://www.AndroidPolice.com Artem Russakovskii

      We don't have any information confirming this but I think it would be silly to assume that one of the newest and most powerful phones will not be supporting the next version.

      There are problems with HTC Sense and JIT that are currently present in custom ROMs but I'm sure HTC has been hard at work making the 2 compatible.

      • Paul

        The CPU on the Hero is a dog, of course the number is lower. The 1 gig snapdragon is the fastest chip out there on a Android phone.

  • michael hill

    How do I get my Droid to get a score like the one on the top 10. Right now I am averaging 4.5. I want 33. Please let me know. Thank u

    • Adam Nabi

      Buy a new phone? How do I get my Fiat Punto to score like a Formula F1? Please let me know. Thank u

  • Noodles

    Did you test to see if the new Nexus One will have a screen reader built in? Could a blind person use it without buying Steel or Standard Android, both of which don't allow users to text, email or surf the web?

  • Sean

    Run the Sprint HTC Hero with an OC'ed 2.1 rom and JIT and you are looking at 4.8 MFLOPS

    • http://www.AndroidPolice.com Artem Russakovskii

      But no Sense, and I like my Sense (they're currently not compatible with each other, I believe).

      • Sean

        Don't get me wrong I love Sense but I am finding way's around it.

        • http://www.AndroidPolice.com Artem Russakovskii

          Oh yeah? Is it noticeably faster? In which areas?

        • Sean

          The boot time is faster, loading and unloading of apps is quicker. I mean it's not like going to a snapdragon cpu but sometimes just that little extra speed helps, like when opening cam to shoot something. The main thing is the OC'ed cpu though. Running nice at 767Mhz.

  • http://garbagecodes.blogspot.com Dody Suria Wijaya

    This is also can be interesting to compare mobile phone processors vs standard desktop/laptop cpu.

    Using linpack at this address:
    http://www.netlib.org/benchmark/linpackjava/LinpackJava.zip

    I got the following score (Intel T7500 2.2 GHz c2d)

    Mflops/s: 228.889 Time: 0.0 secs Norm Res: 1.43 Precision: 2.220446049250313E-16

    or about 32x improvement over NexusOne :D

  • Vidit Bhargava

    So...does wireless-n and FM radio get enabled when you installed Froyo?

  • Dave

    Just ran the test on my N1. Getting 6.3 MFLOPS with everything still running in the background. CANT WAIT for 2.2!!!

    • dave

      ran test on my desire 8.9 first run then 9.2 second run bring on 2.2

      • dave

        sorry it was 6.9 then 7.2

  • John Thompson

    I'd also love to see FM radio - hopefully if it isn't in stock 2.2, cyanogen will include it in his mod in the near future

  • Max

    i'm still waiting for 2.1 on my HTC Hero...sigh

  • Jason Porter

    I'm getting ~4.5 MFLOPS consistently on my Motorola Droid (standard 2.1, Verizon build). I'm very interested to see what JIT would do for this phone.

  • Roach

    But the GUI is still ugly and the apps still suck. It's like having a fast Chevette.

    • Somebody

      No, you're thinking of that fruity phone. This is ANDROID.

  • j maz

    How can you download it?

  • electronic cigarette

    Dude that type of speed is super promising. Bye bye iphone ;)

  • prouddesireowner

    Hi, just read the thread...that's awesome! Did you get a chance to see if the apps2sd issue has been addressed yet? I'd love to see that fixed in froyo!

    --Android rules!--

    • Thomas

      I am also interested about Apps to SD. Thanks

    • Quoth

      According to the issue tracker the status of that particular issue has been changed to 'FutureRelease', which means it's done and just needs to be released (http://code.google.com/p/android/issues/detail?id=1151), presumably in Froyo.

  • Claudio Ibarra

    I don't doubt the photo or the performance increase, but I have a question:

    If the photo reads "37.593" and the Linpack top 10 shows a range of 33.416 to 80.796, shouldn't one of the entry rows read "37.593" ?

    • http://www.AndroidPolice.com Artem Russakovskii

      Only if you specifically save it to their online db.

  • Fábio

    I got 6.8 MFLOPS on my Android2.1/HTC Desire.
    Can't wait for the update!!!! ^_^

  • Fábio

    I got 6.9 MFPS on my HTC Desire / Android 2.1.
    And wow can't wait for the 2.2 update

  • polkadots

    I cannot help but notice that the Droid did pretty darn well at about 34 MFLOPs for a 550+Mhz processor. I am using a Milestone but I reckon it should get about the same performance boost. Froyo!! Can't wait!!

  • Joshua

    So is the the thing that one of the exiting programmers said would "blow us away". I mean it's great and everything but i'm not sure it really has the "blow me away" factor considering that i really like the speed of my nexus one to begin with. I'm really interested in what's gonna be the best improvement

  • jasim

    does any one know if it include arabic support?

  • http://www.fromdev.com Priya

    That sounds weird. Anyways, now m interested to know if its faster then iPhone 4?

    • Tetsuo

      I am not sure if Iphone 4 has a similar benchmarking tool, it would be nice to do a comparison, but Froyo is breathing new life to my old HTC Hero.. ;) It is clocking 4.17 MFlops on it running a custom rom ;)

  • RoyBatty

    I just found another benchmark for Java performance on Android: caffeinemark. The great thing is that you can see how Froyo improves the results for a wide range of devices/ROMs ... on this website :

    http://www.flexycore.com/benchmark-database-access.html

  • http://www.itrush.com IT Rush

    Amazing, maybe I'll consider updating to this cool device asap.. Looks very interesting really!

    • Dextroz

      Are you kidding me? This post is from a year ago! The Nexus One is now rocking Gingerbread 2.3.

      Do yourself a favor and get a Nexus S at this point instead.

  • http://www.itrush.com IT Rush

    Wow, what an improvement.. can't wait to get one. Looks like it's time to drop my other phone and switch to this one eh.

  • kia

    nice page I have learn a lot from this page very nice

  • http://kids-quran-online.blogspot.com/ kristen

    well its is faster then i.4 looks to me and but i.5 steels the race

  • http://hivestudition.com Jasa animasi flash

    Its fastest than iphone 4 wow its great

  • http://tutiez.com/enum-in-java-with-java-enum-examples.html pranav

    LINPACK benchmark helps a lot .. indeed

  • xiaoying

    <div>Windows 7 tidlig vedtagelse beats Vista's 2 til 1</div>
    <div>I fem måneder siden dens udgivelse, har Windows 7 erobrede to gange de forbrug andel som  levis 501
    sin forgænger, Vista, nås i den samme tidsperiode, et Web metrics selskab sagde i dag.Ifølge NetApplications.com i Aliso Viejo, Californien, Windows 7 tegnede sig for 9% af alle levis jeans uk
     operativsystemer i levis commuter jeans 
     brug online for måned til sidste søndag. Til sammenligning havde Windows Vista 4,5% andel fem måneder efter dens  levis 501 jeans
    sene-januar 2007 frigivelse til detail.Microsoft Corp. leveret Windows 7 22. oktober, 2009.Windows 7's bane har været hurtigere end Vista's nogensinde siden dens udgivelse; den nyere operativsystem ramt 4% mærket flere måneder tidligere end Vista gjorde. Ved udgangen af januar var Windows 7 forbrug andel 7,5%, også dobbelt på 3,75%, Vista nydt ved udgangen af den fjerde måned".Ser man på tendenserne, [Windows 7] væksten synes at være stærk og konsekvent, med ingen synlige nedgang,"sagde Vince Vizzaccaro, executive vice president på NetApplications.Vizzaccaro bemærkede også, at forskellen i Windows 7 weekend og ugedag scores har været stigende, et tegn på, at"personlige forbrug vokser hurtigere end virksomhedens forbrug, som passer til forventningerne,"sagde han. Tidligere har Vizzaccaro forklaret, at forbrug andelen af nyere versioner af Windows klatrer i weekenden, når en større procentdel af computere online er hjem maskiner. Virksomheder halter traditionelt bagefter hjemmebrugere i udbredelse af nye versioner af Windows.Microsoft's operativsystemer tegnede sig for en kombineret 92% af alle operativsystemer kraftoverførsel computere, gik online sidste måned. Windows XP udgjorde hovedparten af, med en markedsandel på 66.3%, mens Vista kørte på 17,4% af maskinerne i NetApplications undersøgelse. Apple Inc.'s Mac OS X havde en andel på blot 5,1%.Hvis tendenserne i de sidste tre måneder fortsætter, Windows 7 vil overhale Vista som den anden-de fleste-populære operativsystem, bag No. 1 XP, ved June.NetApplications foranstaltninger operating system forbrug ved at spore maskiner at besøger 40.000 overvåger for klienter, som resulterer i en pulje af ca. 160 millioner unikke besøgende pr. måned.   Windows 7 forbrug andel er dobbelt Windows Vista's gennem hver edition's første fem måneder. Gregg Keizer dækker Microsoft, sikkerhedsproblemer, Apple, Web-browsere og generelle teknologi bryde nyheder til Computerworld. Følg Gregg på kvidre ved @ gkeizer, sende en e-mail til gkeizer@ix.netcom.com eller abonnere Gregg's RSS feed.</div>