The Kindle Fire's Silk Browser

A couple of posts ago I promised a review of the Kindle Fire coming soon, and I still plan to write that review. In fact, I plan to basically write THREE reviews: 1) The Kindle Fire as a personal media consumption device (for which it was intended), 2) The Kindle Fire as a productivity tool for educators (taking the place of--say--a Smart phone or an iPad, and 3) The Kindle Fire as an instructional tool for use with students. Before I do that, though, I want to write about one part of the Kindle Fire that I am truly disappointed about, and that's the Amazon Silk browser.

When the Kindle Fire was announced several weeks ago (and by the way, I bought my Fire less than an hour after it went on sale that day), one of the items promoted heavily was the web browser built into the Fire. It was called "Silk," and it was going to revolutionize the Internet web browser experience by taking the processing power out of the hands of the local machine and putting it in the hands to Amazon's massive cloud of servers. It was SUPPOSED to speed up web browsing on the Kindle Fire significantly.

But then, once the Fire was released last week, reports started coming in that the Silk browser was pretty slow, and that it actually was SLOWER when the Acceleration was running. I didn't believe it, and so I set out to see for myself. And what I found was...all of those websites were right. The browser is slower when the Accelerator is running!

To test this theory, I loaded two websites (THIS website and http://cincinnati.com, a site which--in my experience--takes a VERY long time to load) with 1) my home computer (2.4 Ghz Core 2 Duo Windows 7 machine with 3 GB of RAM) and with the Fire both 2) with the accelerator turned on and 3) off. I loaded each site  five different times for each, clearing the history and cache of all three after each site visit. I then used a stop watch to see how long it took each site to fully load. Here's what I found:

Site: Sweasy.net

Attempt	Computer	Fire w/ Acceleration	Fire w/o Acceleration
1	5.9	9.2	7.3
2	3.6	14.4	8.0
3	3.7	7.4	6.3
4	3.7	11.7	7.4
5	3.7	7.4	8.2
Average	4.12	10.02	7.44

Site: Cincinnati.com

Attempt	Computer	Fire w/ Acceleration	Fire w/o Acceleration
1	4.5	11.9	12.4
2	4.9	12.8	12.1
3	4.4	11.6	11.4
4	4.1	12.3	10.9
5	3.7	11.1	10.7
	4.32	11.94	11.5

I would describe this website as taking a moderate amount of time to load (It has a Twitter feed and an eschool news feed which slow it down, as well on this day as a YouTube video on the home page which took a few second to load). After five tries, my computer loaded the page in an average of 4.12 seconds. The Kindle Fire without acceleration turned on loaded the page in 7.44 seconds, taking almost twice as long. When the accelerator was turned on, though, the Kindle Fire took more than 10 seconds to load! That's much more than twice as long as it took my computer to load, and it's 25% longer than it took the Fire to load in its "slower" mode with the accelerator turned off.

When I ran the same test at Cincinnati.com the results were a little less dramatic. There's a LOT of loading of secondary content on that site (ads and video and external news feeds), and while loading the page on my computer only took a few tenths of a second longer than this web site did, both the Silk accelerated and Silk non-accelerated tests took more than 11 seconds. The accelerated browsing still took a little longer, but this time was only .44 seconds slower than the Fire was when it had acceleration turned off.

But I guess that's missing the point. The browser is supposed to be FASTER when the accelerator is turned on, not "just a little bit slower." So regardless of whether or not the Amazon Kindle Fire is a success, I'd have to call their Silk web browser a failure. A big one.