To hear Amazon tell it, the Amazon Silk browser is going to be a revolution in how mobile devices browse the web. Silk acts as an intermediary between your mobile device (the Kindle Fire) and the Internet at large. As you make a request -- go to TheAwl.com -- Silk jumps in and distributes the workload of finding, downloading, and presenting that information on the device. If you use your phone or current tablet to do that stuff, there are a bunch of servers that your gadget has to talk with in order to get the data it needs to render a page. Each step in that process, as Brett Taylor, the principal product manager for Silk explains in the video above, takes time.
"With every page request... you gotta go out and do some DNS resolution, figure out where's the origin server, issue that request, do sort of a TCP handshake, ask for the content that you want, get the acknowledgement back to you, bunch of back and forth steps to kind of boil it all down," Taylor says. "Your device is going to have go back and forth across that wireless network perhaps dozens of times. And that's all time."
With Silk, Amazon's cloud computing infrastructure takes care of a lot of those steps and serves you up just what you need. That seems like a very good idea and will reduce the strain on the Fire's resources.
Two other features of the browser are really interesting. The first is automatic content scaling. So if a webpage has a huge 3MB JPG embedded, Silk will make it smaller knowing that you won't be able to tell the difference on your mobile device. Smart, if it works. The second is predictive caching a la Google Chrome. Amazon tries to take the aggregate user behavior and preload their best guess of where you might click next.
It all sounds awesome, but I'm reserving judgment. Here's why:
1. The Opera browser has been doing a lot of this stuff for a while now and it has a two percent market share. Update: @allyngibson points out that Opera Mini has a 22 percent share, quite possibly destroying my point here.
2. Google Chrome's integration with Google Search allows them to predictively load content, too. It seems to make a big difference in demos, but in my real life, I could barely notice the difference.
3. Amazon didn't release any metrics about how much faster Silk could load web pages than its competitors. Not one number!
4. The real limiting factor for mobile browsing is still connectivity and bandwidth. It's hard to find a good 3G network connection way outside of the cities and the ones inside the cities are clogged. No matter how many neat tricks Amazon uses to make up for our crappy, overloaded mobile Internet infrastructure, the crappy, overloaded mobile Internet infrastructure remains. Update: I got ahead of myself here, imagining the Kindle Fire 3G that I'm sure will come down the pike. For now the Kindle Fire is Wi-Fi only as @andreasudo points out.
So, hats off to Amazon for trying to jerryrig systems to make the mobile web go faster. I just wonder if we'll actually notice the difference given all the variables that are outside of the company's control. One last thing to watch: Silk's backend has been designed to learn from its users' behavior. Even if the service isn't great at launch, it may get better as Amazon learns what pages it should have ready to serve up to people.
We want to hear what you think about this article. Submit a letter to the editor or write to letters@theatlantic.com.
