Amazon's Silk does little to accelerate Kindle Fire, HTML5

Posted:
in iPad edited January 2014
An assessment of the new Amazon Kindle Fire's capabilities a web app client found it a "competent but minimal HTML5 platform" but reports that a key feature, its Silk server-side accelerated browsing is "not very noticeable."



HTML5 development tools vendor Sencha put the Fire through a series of tests, as it has previously done with Apple's iPad 2 and iOS 5, Android 2.x tablets like the Galaxy Tab, Android 3.x Honeycomb tablets like the Motorola Xoom, and RIM's BlackBerry PlayBook.



The firm notes that "the Fire runs a customized version of Android 2.3.4 (Gingerbread) and in the past, we?ve been disappointed with the quality and completeness of that browser. We were hoping that Amazon would improve the stock Gingerbread browser significantly."



It concluded "we can say that while it?s a solid browser for normal page browsing, it lags the best of the competition in HTML5 capabilities. Constrained by its Gingerbread foundation, it?s a competent but minimal HTML5 platform that reflects its $200 price point."



Web standards testing



In terms of web standards testing, the Fire scored 95 out of 100 in the Acid3 test, below both iPad 2 and the Playbook, both of which soured 100. It lost points primarily due to lacking support for SVG, failing to complete media queries tests in the allotted time, and being too slow to pass a "challenging" garbage collection speed test.



Modernizr, a test that uncovers browser support for new web features often associated with HTML5, shows the Fire browser supports CSS 2D transforms and Canvas "as well as other Android 2.x capabilities."



Fire was also found to correctly render both Google Fonts and Typekit dynamic fonts, although the reports notes that both Typekit font loading and page scrolling with lots of Google Fonts were both "noticeably slow."



However, the Fire's year old version of Android does't support 3D transforms, web sockets, web workers and many HTML form input types, and the Fire's minimal hardware lacks support for web features such as geolocation.



SunSpider and web animation performance testing



In terms of SunSpider performance, a test that benchmarks JavaScript, the Fire and its dual core chip (the same as the PlayBook) scored "in the same range as other tablets," an impressive feat for a device that costs less than half as much.



In real world testing of CSS3 animation performance however, Sencha reports that while a test game did play, "the frame rate was poor and touch responsiveness while animations were running was also substandard."



The report notes that the browser didn't seem to use the Fire's built in GPU core, in contrast to the similar PlayBook, which "does a far better job leveraging the same GPU core."



In simpler animation testing, Sencha notes that the Fire's browser allowed animation elements to "diverge and lag noticeably and visibly from each other," an issue it blamed on Android's JavaScript timer latency, which it contrasted with iOS 5's "best-in-class implementation."



Out of touch performance



In testing general performance as an HTML5 web app client, Sencha noted that "the Kindle Fire has problems processing touch events with good responsiveness," noting that like previous Android devices, "the OS and browser seem to fight over who gets touch events."



The report added, "since the Fire is based on Android 2.x, full multi-touch with independently tracked touches is not supported either."







Silk has little impact



"One of the main selling points of the Kindle browser is supposed to be its cloud-caching and pipelined HTTP connection that uses the SPDY protocol," Sencha wrote. "This does seem to speed up normal page browsing a little, but it?s not very noticeable and we didn?t test this rigorously. But for HTML5 web apps, where code is downloaded and executed, there doesn?t seem to be any performance difference when we tested with acceleration on and off."



The company also notes that "SunSpider scores were essentially the same when accelerated browsing was turned on or off."



The report concludes, "Amazon Kindle Fire doesn?t seem designed to run HTML5 apps as a primary goal. It does a good job of displaying ordinary web pages and its resolution and rendering capabilities meet that need well. But there are too many sharp edges, performance issues, and missing HTML5 features for us to recommend that any developer create web apps primarily for the Kindle Fire. The iPad 2 running iOS 5 continues to be the tablet to beat, with the PlayBook a respectable runner-up in HTML5 capabilities."



Amazon Kindle Fire browser testing from Sencha on Vimeo.



Sencha has also posted a video that demonstrates the minimal acceleration Silk seems to provide when loading a variety of complex and simple websites, which in some cases actually slows down rendering as much as it speeds it up in others.
«134

Comments

  • Reply 1 of 76
    I expect some of these shortcomings to be fixed when Amazon updates the Kindle Fire OS to Android 4.0. Other than that, I wasn't really surprised in the results. Android 2.x isn't that great when it comes to supporting modern standards. Performance is subpar and only masked on phones because of processors that range from 1.2-1.5Ghz to make up for the OS' shortcomings.



    Considering Apple gets better performance only running it's CPUs at 800Mhz, it's pretty clear to me that software matters a LOT. Heck, Windows Phone 7 devices are smoother than most Android smartphones and they only have a single core (but hardware acceleration!).



    It's the software, stupid!
  • Reply 2 of 76
    29922992 Posts: 202member
    fair enough!
  • Reply 3 of 76
    solipsismsolipsism Posts: 25,726member
    Those that are buying this as a tablet aren't likely to care much. It's the same mentality that has those buying $400 PCs and still running IE6. It's just not something they consider.





    PS: Speaking of JS performance has Google dropped the ball with their engine or just dropped the ball with their Android browser? As you can see in the link below the Galaxy Nexus and Droid RAZR on par with the iPhone 4S in JS yet those machines have a 50% faster CPU and 2x more RAM. I'd think they should be doing much better in JS performance as Chrome browser is over Safari on the same HW. And pushing almost 2x as many pixels as the iPhone 4/4S but using a considerably weaker GPU? Does that really make sense to any Android buyer?
  • Reply 4 of 76
    Quote:
    Originally Posted by dagamer34 View Post


    I expect some of these shortcomings to be fixed when Amazon updates the Kindle Fire OS to Android 4.0.



    Why would Amazon upgrade the Fire to Android 4.0?



    3.0/4.0 are Google's efforts to make Android into an iPad challenger.



    Kindle Fire is Amazon using the Android 2.x installed base to instantly have a bunch of apps the Fire can run. "Upgrading" to 3/4 would force Amazon (or B&N) to adopt Google's direction, rather than go on their own as device makers taking advantage of the fact that more than half of Google's own app market customers are still running 2.2.



    Amazon didn't update its original Kindles aggressively, and no Android 2.x vendor has upgraded its 2.x tablets to 3.x.



    Further, even if Amazon could pick and choose portions of newer Android software to use on future Fire devices, Android's browser is not nearly as good as iOS 5's in HTML5 performance. If anything, Amazon is helping make Android more of a mess than it already is.



    If the Fire takes off, it will put a bullet in the head of Android 4.x tablets, and they've already shot themselves full of lead. Will be interesting to see how well Google maintains any leadership when "Android" on tablets comes to mean $200 devices purposely running last years 2.3 version.



    If Google can't give away its latest version of Android, it becomes more pathetic than even Microsoft and its inability to sell Windows Phone.
  • Reply 5 of 76
    Thanks for all the great Kindle coverage.
  • Reply 6 of 76
    apple ][apple ][ Posts: 9,233member
    I saw some of the reviews for the Kindle Fire and it's precisely how I thought it would be, complete fucking garbage. Did anybody expect anything different from an Android tablet? What experience does Amazon have with making hardware and software, besides their kindle readers?



    Some reviewers said that it doesn't even qualify to be called a tablet.



    The Silk browser is slow as shit. I guess besides datamining people's surfing habits, there's not much benefit to having a browser in the cloud, because it sucks. The tablet is also too small for browsing, and when you go to zoom, it's choppy and slow as hell. It's also too small for comics, magazines and other reading purposes.



    The general UI of the tablet is laggy, unresponsive and often doesn't register clicks. Oh, it's not multitouch either, it's only two touch and the one gesture it has, zoom, fucking sucks because of the choppiness, like I already mentioned above.



    And the power button is on the bottom? What kind of moron thought up that idea? I read that quite a few reviewers were questioning that retarded move. It doesn't even have physical volume buttons. That's accessed through the crappy touch screen.



    Some people say that this tablet is like an upgraded Kindle. No it's not. An e-ink Kindle is a better reader, so what exactly is this Amazon thing good for? Not fucking much.



    I'd rather not have any tablet at all, than to have this Amazon tablet. The $200 price tag doesn't mean shit. It may be half the size of an iPad, and less than half the price of an iPad, but the quality is about 10% of an iPad. It's unacceptable. If I picked up this tablet, I would probably end up throwing it into the wall after a few minutes of using it. Slowness, lagginess, and choppiness = Death.



    Oh and one last thing. That Amazon one-click purchasing is sure going to put a hurt on a lot of poor people, the people who are buying this tablet to begin with. If people have no money to buy something proper, and they settle for junk, even if it costs $200, then these people certainly don't have the money to be buying much from Amazon. Letting their kids use the tablet for two minutes can end up being a very costly thing for these poor ignorant people who make misguided and uninformed choices.
  • Reply 7 of 76
    drdoppiodrdoppio Posts: 1,132member
    Quote:
    Originally Posted by Suddenly Newton View Post


    Thanks for all the great Kindle coverage.



    Indeed. Such an insignificant gadget getting so much attention!



    Quote:
    Originally Posted by Corrections View Post


    Why would Amazon upgrade the Fire to Android 4.0?

    ...



    Amazon didn't update its original Kindles aggressively, and no Android 2.x vendor has upgraded its 2.x tablets to 3.x.

    ...



    It seems from the poor reviews that they will have to update now. We shall see...



    Quote:
    Originally Posted by solipsism View Post


    Those that are buying this as a tablet aren't likely to care much. It's the same mentality that has those buying $400 PCs and still running IE6. It's just not something they consider.



    Exactly.



    Quote:

    PS: Speaking of JS performance has Google dropped the ball with their engine or just dropped the ball with their Android browser? As you can see in the link below the Galaxy Nexus and Droid RAZR on par with the iPhone 4S in JS yet those machines have a 50% faster CPU and 2x more RAM. I'd think they should be doing much better in JS performance as Chrome browser is over Safari on the same HW. And pushing almost 2x as many pixels as the iPhone 4/4S but using a considerably weaker GPU? Does that really make sense to any Android buyer?



    Thanks for the link. I don't quite understand your concerns though. The Nexus and RAZR both outperform the 4GS by about as much as their CPUs are faster, and RAM isn't of importance here. As to GPU performance, they are behind the 4GS but 3 times faster than the 4G. I wouldn't question their graphic performance unless there have been any serious issues observed with the three times slower 4G. It actually seems that the performance boost of the 4GS over the 4G is a bit of an overkill, unless it can do 9 times more in real-life applications.
  • Reply 8 of 76
    Silk is an excellent feature, but I'm not surprised it doesn't give a big advantage over WiFi...

    However, should Amazon release a 3G version of the Fire... Well, then you will really notice a difference with Silk!



    The same things happens to BlackBerries. Browsing in 2G (and even in 3G, sometimes) is much faster on BlackBerries than in other phones, thanks to the BIS...
  • Reply 9 of 76
    Quote:
    Originally Posted by Apple ][ View Post


    It may be half the size of an iPad, and less than half the price of an iPad, but the quality is about 10% of an iPad. It's unacceptable. If I picked up this tablet, I would probably end up throwing it into the wall after a few minutes of using it. Slowness, lagginess, and choppiness = Death.



    Finally, a Fire review that I can agree with.
  • Reply 10 of 76
    Quote:
    Originally Posted by EgoAleSum View Post


    However, should Amazon release a 3G version of the Fire... Well, then you will really notice a difference with Silk!



    Yeah, sure, it should be included in the $200 price.
  • Reply 11 of 76
    solipsismsolipsism Posts: 25,726member
    Quote:
    Originally Posted by DrDoppio View Post


    The Nexus and RAZR both outperform the 4GS[sic] by about as much as their CPUs are faster



    They don't. In one JS test the Droid RAZR even did slightly worse than the iPhone 4S despite having a clock speed 50% faster than in the iPhone. And in no test was the these 1.2GHz devices showing 50% faster JS rendering. What is the point of having a 50% faster CPU over optimized code?
  • Reply 12 of 76
    Silk is just an excuse for data collection. Anyone with a basic understanding of networking would realize that the performance boost would be unnoticeable, and it might actually make things worse in many conditions.
  • Reply 13 of 76
    sflocalsflocal Posts: 6,092member
    Quote:
    Originally Posted by solipsism View Post


    They don't. In one JS test the Droid RAZR even did slightly worse than the iPhone 4S despite having a clock speed 50% faster than in the iPhone. And in no test was the these 1.2GHz devices showing 50% faster JS rendering. What is the point of having a 50% faster CPU over optimized code?



    Solips, you and I know that efficiency will always trump raw performance. Like Windows, Android is just a patchwork that needs higher-horspower specs to do barely the same functions that iOS can do.



    Makes me wonder if tech-heads that continue to get wet on specs are simply ignorant, or just pretend to know what they are talking about.
  • Reply 14 of 76
    dave k.dave k. Posts: 1,306member
    Quote:
    Originally Posted by bedouin View Post


    Silk is just an excuse for data collection. Anyone with a basic understanding of networking would realize that the performance boost would be unnoticeable, and it might actually make things worse in many conditions.



    I call BS. I have a networking background, and I think that the service has potential... Amazon just started selling their Kindle Fire's. Perhaps not everything is working just yet.
  • Reply 15 of 76
    gatorguygatorguy Posts: 24,176member
    Quote:
    Originally Posted by solipsism View Post


    They don't. In one JS test the Droid RAZR even did slightly worse than the iPhone 4S despite having a clock speed 50% faster than in the iPhone. And in no test was the these 1.2GHz devices showing 50% faster JS rendering. What is the point of having a 50% faster CPU over optimized code?



    I'm confused why you're trying to move the discussion into the Nexus and RAZR, neither of which have anything at all to do with Amazon's Fire. They don't share hardware, OS versions or markets



    If you wanted to open a thread to compare them I would imagine it might get some responses, but this really isn't the proper one is it? Kinda derails the discussion.
  • Reply 16 of 76
    postulantpostulant Posts: 1,272member
    Quote:
    Originally Posted by solipsism View Post


    What is the point of having a 50% faster CPU over optimized code?



    Because optimized code requires effort and a belief that software and hardware should form a marriage within a product. It's much easier to slap in a faster CPU and hope that the lousy software optimization will just go unnoticed.



    When your software is shit, you're left with little choice but to focus on the hardware.
  • Reply 17 of 76
    apple ][apple ][ Posts: 9,233member
    Quote:
    Originally Posted by Gatorguy View Post


    I'm confused why you're trying to move the discussion into the Nexus and RAZR, neither of which have anything at all to do with Amazon's Fire. They don't share hardware, OS versions or markets



    If you wanted to open a thread to compare them I would imagine it might get some responses, but this really isn't the proper one is it? Kinda derails the discussion.



    You're talking to the wrong guy. He wasn't the one that brought it up.
  • Reply 18 of 76
    apple ][apple ][ Posts: 9,233member
    Quote:
    Originally Posted by anantksundaram View Post


    Finally, a Fire review that I can agree with.



    Much obliged.
  • Reply 19 of 76
    Quote:
    Originally Posted by solipsism View Post


    They don't. In one JS test the Droid RAZR even did slightly worse than the iPhone 4S despite having a clock speed 50% faster than in the iPhone. And in no test was the these 1.2GHz devices showing 50% faster JS rendering. What is the point of having a 50% faster CPU over optimized code?



    I checked out the Android 4.0.1 source and had a look. They're using v8 version 3.2.10, from October 20, 2011 - http://code.google.com/p/v8/source/detail?r=9719. For comparison, the current stable version of Chrome (15.0.874.121) uses v8 version 3.5.10.24, which has several performance improvements. Android 4.0.1 isn't yet using the NEON-optimized support libraries, such as libjpeg-turbo (http://libjpeg-turbo.virtualgl.org/). I'm not sure why this is, but I suspect Google will upgrade their external libraries in a future version.



    Of course, when they do that, I'm sure the fair-minded folks here will be sure to reevaluate their opinion of Android based on the evidence, right?
  • Reply 20 of 76
    postulantpostulant Posts: 1,272member
    Quote:
    Originally Posted by sflocal View Post




    Makes me wonder if tech-heads that continue to get wet on specs are simply ignorant, or just pretend to know what they are talking about.



    It's what they've been taught - That specs tell the story. Pick up Best Buy's sales ad tomorrow and see how they market PCs and Android phones: Ghz, GB, Robots, and explosion.
Sign In or Register to comment.