As reported earlier Tuesday, Amazon has released a software update for its Kindle Fire tablet, promising enhanced â€œfluidity and performanceâ€ and improved â€œtouch navigation responsiveness.â€ But nowhere in its update language does Amazon speak to improvements in the speed of its Silk browser.
As it turns out, the new software update doesnâ€™t seem to improve browser performance to a significant degree. And thatâ€™s a shame, because laggy web-page load times are a real problem for Amazonâ€™s tablet.
When I reviewed the Fire a little less than five weeks ago, I found the tabletâ€™s browser performance to be bizarrely, inexplicably slow. The Fire, after all, has a 1GHz dual-core processor, just like the iPad 2 and all the Android/Honeycomb competitors. So on processing power alone, the Fire should have all the hardware it needs to deliver fast web browsing.
Whatâ€™s more, the Fire is also supported by Amazonâ€™s much-ballyhooed Silk technology, which splits processing and data-fetching workloads between the tablet itself and the cloud. As Amazon CEO Jeff Bezos stated in a Sept. 28 press release before the Fireâ€™s launch, â€œWe refactored and rebuilt the browser software stack and now push pieces of the computation into the [Amazon Web Services] cloud. When you use Silk â€” without thinking about it or doing anything explicit â€” youâ€™re calling on the raw computational horsepower of Amazon Elastic Compute Cloud to accelerate your web browsing.â€
On paper, Silk web browsing sounds marvelous. But when I tested the Kindle Fire in November, I found page load times to be double and sometimes triple that of an iPad 2.
So, now, some five weeks later and updated for improved performance, how does the Kindle Fire stand up? After all, itâ€™s not only received a software update to improve â€œfluidity and performance,â€ but the Silk browsing architecture has theoretically had time to â€œlearnâ€ the browsing behaviors of all Kindle Fire owners.
As Amazon describes it, â€œTraditional browsers must wait to receive the HTML file in order to begin downloading the other page assets. Silk is different because it learns these page characteristics automatically by aggregating the results of millions of page loads and maintaining this knowledge on EC2 [Elastic Compute Cloud].â€
Iâ€™m sorry to report that even after updating my tablet to the 6.2.1 OS build, Fire page loads still lag significantly behind the very same loads on iPad 2.
Yes, it appears the Fireâ€™s browsing performance has improved, and during testing I didnâ€™t see 300 percent performance gaps between the Fire and iPad 2. But the numbers Iâ€™ve collected still indicate the Fire isnâ€™t delivering on Amazonâ€™s silky promises.
Below are some specific page load comparisons. Before testing anything, I cleared each tabletâ€™s browser cache and history. For each test, I loaded a siteâ€™s URL into the browserâ€™s address bar, and used a stopwatch to measure the time between hitting â€œgoâ€ on the onscreen keyboard, and when the very last object loaded in the page. Also, for the Kindle Fire, I left on the system default â€œAccelerate Page loading.â€ This toggle allows the Silk browser to tap into Amazonâ€™s back-end cloud servers. Load times are measured in seconds; lower scores are better.
Wired.com â€” Fire: 7.1; iPad 2: 5.2
NFL.com â€” Fire: 13.6; iPad 2: 11.0
NBA.com â€” Fire: 13.6; iPad 2: 5.3
Microsoft.com â€” Fire: 5.4; iPad 2: 2.6
TheVerge.com â€” Fire: 13.4; iPad 2: 11.1
BoingBoing.net â€” Fire: 20.0; iPad 2: 13.8
Imdb.com â€” Fire: 9.8; iPad 2: 5.3
BBC.co.uk â€” Fire: 8.0; iPad 2: 6.6
Yahoo.com â€” Fire: 5.8; iPad 2: 2.7
Amazon.com â€” Fire: 8.5; iPad 2: 4.6
Iâ€™ll be the first to concede my testing lacks a number of important controls. While I tested both tablets just two feet away from a Wi-Fi router, I didnâ€™t test each page load simultaneously, but rather sequentially â€” and varying server loads on the content-provider end can affect browser load times. And, of course, I used a simple stopwatch, not sensitive testing equipment, to measure page load times.
Nonetheless, my testing backs up my own anecdotal user experience: Web browsing on the Kindle Fire is still palpably slower than on the iPad 2. Iâ€™m no longer seeing page loads that take three times longer, but the performance delta still ranges from â€œnoticeableâ€ to some 200 percent.
Fire apologists will try to â€œexplain awayâ€ the poor performance by reminding critics that the tablet is still only $200, a veritable impulse buy. And even I will concede that the Fire offers a nice set of features for its low entry price. Nonetheless, Amazon has celebrated Silk as breakthrough technology that supercharges browser performance, and nothing about the Kindle Fire experience delivers on that promise.
Posted by Janine E. Mooney, Editor
December 22, 2011