Kindle Fire

Earlier reports of the device now known as the Kindle Fire have varied wildly and with speculation rampant about what Amazon might announce the finished product appears to fall nicely in between the greatest device ever and a serious disappointment. We'll begin with pricing. Jeff Bezos, Amazon CEO, repeated one line more than any other during the event today, "premium products at non-premium prices." By pegging the Fire at $199 he certainly is following through on the latter claim. This undercuts even the Barnes and Noble Nook Color, while providing specifications that match devices more than twice it's price. So, bargain? You bet.


And what you get for that $199 is a stylish black device with a 7" IPS 1024 x 600 screen, and a 1 GHz dual-core ARM processor. At 11.4mm the Fire is thicker than other slates, but as we've discussed before, a thicker device can still be pleasant to hold, so long as the form factor works. Reports from gdgt's Ryan Block indicate that Quanta Computers, who designed the BlackBerry PlayBook, were responsible for the design of the Kindle Fire, and by all accounts they seem to have not strayed far from that design. Holding to that design may include using the same TI OMAP 4430 SoC, though we have yet to confirm that. What we can confirm is that at 413 grams, this is one of the lightest weight tablets we've seen. 

Tablet Specification Comparison
  Amazon Kindle Fire Apple iPad 2 BlackBerry PlayBook Samsung Galaxy Tab 8.9
Dimensions 190 x 120 x 11.4mm 241.2 x 185.7 x 8.8mm 194 x 130 x 10mm 230.9 x 157.8 x 8.6mm
Display 7-inch 1024 x 600 IPS 9.7-inch 1024 x 768 IPS 7-inch 1024 x 600 8.9-inch 1280 x 800 PLS
Weight 413g 601g 425g 447g
Processor 1GHz TI OMAP 4430 (2 x Cortex A9) 1GHz Apple A5 (2 x Cortex A9) 1GHz TI OMAP 4430 (2 x Cortex A9) 1GHz NVIDIA Tegra 2 (2 x Cortex A9)
Memory ? 512MB 1GB 1GB
Storage 8GB 16GB 16GB 16GB
Pricing $199 $499 $499 $469

Hardware isn't the whole store with the Fire, though. Amazon is selling a platform from which to experience it's various Amazon services, and it is leaving none of them from this party. Obviously, and this is the last mention of Android you'll find in this piece and their PR. For third party apps there's the Amazon Appstore. This will be the only official means by which buyers will be able to load their apps onto the device, though intrepid hackers will no doubt make quick work of the device. Amazon's MP3 store is on hand as well as Amazon's Kindle app and e-book store. This brings us up to apps, music, magazines and books. The addition of Amazon Prime Instant Videos rounds out the offerings with over 100,000 movies and TV shows from the likes of Fox, CBS and NBCUniversal. Buyers will be treated with an expanded WhisperSync service that now allows users to mark their place in movies and TV shows, as they already do with books and magazines. Amazon is even leveraging EC2, their web services provider to enhance the browsing experience with Amazon Silk.

Amazon Silk Browser

The Kindle Fire is the first device to ship with Amazon's Silk browser. According to Amazon, Silk is a split-browser that leverages the company's EC2 cloud. The idea is simple: rather than simply going out and maintaining connections with all of the various servers that supply content for a single web page, Amazon will deliver as much of that as possible via its own cloud. The benefit of delivering content via Amazon's cloud is it minimizes the number of independent DNS lookups and handshakes the browser has to perform to load a single webpage, not to mention that Amazon's cloud should hopefully deliver more consistent performance due to the scale of its deployment.
Amazon is also doing some prediction in Silk. As Amazon's EC2 cloud aggregates usage data from Silk users, it can develop models for web page access patterns. The browser is then capable of prefetching what it thinks will be the next web page you click on based on historical data from a number of other users. I've always thought browser prefetching makes a lot of sense and it appears that Amazon is going to try to do some of that here with Silk.
Amazon's cloud can also deliver size optimized content depending on the device you're using. The example Amazon gives is a 3MB jpg that can be compressed to 50KB without you being able to tell the difference on a Kindle Fire. Silk will work with EC2 to determine the appropriate size of the image that it should send down to you. 
Given that the Kindle Fire is WiFi only, I wonder how much of an impact Silk's network optimizations are really going to have on the overall experience - particularly on fast WiFi connections. Web browsing on mobile devices is still largely CPU bound and with the Kindle Fire shipping with (presumably) the same CPU power as a number of other Android tablets that bottleneck shouldn't really change because of Silk.
We also don't know anything about Silk's underlying parsing and rendering engines. As Google has shown us, web page rendering performance depends just as much on software as it does hardware. Amazon is addressing the network part of the problem with Silk but there are also CPU and software issues that are either unchanged or big unknowns at this point.
A $79 Kindle! And Wrap-up
Comments Locked


View All Comments

  • JasonInofuentes - Wednesday, September 28, 2011 - link

    If it is the OMAP 4430 then it'll be as well performing at gaming as any other tablet out there, bar the iPad.
  • slashbinslashbash - Wednesday, September 28, 2011 - link

    So instead of web browsing using up only the user's bandwidth, it also uses up Amazon's bandwidth.... twice (once inbound, once outbound, although they will probably implement caching to reduce some of the inbound). As mentioned in the article, download speed does not really seem to be the bottleneck when it comes to Web browsing on tablets (although CPU+GPU time will be less eaten up by smaller graphics that don't need resizing to display). The biggest problem is that this system will require a long-term dedication of resources on Amazon's part. Same with the free lifetime 3G service on some Kindles.... Amazon sells you the Kindle once but is required to maintain services for it indefinitely. This is not a good business model from Amazon (and of course, at $200 they can hardly be making anything from the hardware, if they're not actually losing money on it). Amazon has awesome sales numbers but barely eeks out a profit from its operations. It is amazingly well-run, but retailing is always a race to the bottom on pricing. This tablet doesn't seem like it will help much; most everybody who wants an e-reader already has a normal Kindle, so the Kindle Fire won't add many incremental e-book sales. Android apps could turn into a profit center, but they have a long way to go before their sales catch up with Apple.

    All that said, as a long-term Apple fan, user and shareholder, this tablet does concern me to a certain degree. 7" is half the screen area of 10", and there are other apparent flaws, but selling at a 60% discount to the cheapest iPad can make up for a whole lot of flaws. I think that Apple will have to drop pricing to compete, maybe by keeping the iPad 2 in the lineup at a reduced price when they introduce the iPad 3. Other Android tablets have failed because of price and marketing reach. The Nook Color has been limited mostly by B&N's flawed marketing (if they sold it as a full-fledged Android tablet instead of a colorful e-reader, I think it would do better). Amazon has both problems fixed, and also has the value-add of their own curated App Store. I think that the Kindle Fire will easily take the 2nd place position in tablet sales.

    This creates an interesting dynamic between Amazon and Google. Amazon is basically forking Android with this tablet. Google can't be happy if Amazon takes over the #1 Android Tablet position with a non-official Android tablet. I would personally be pleased if Amazon purchased WebOS from HP and put it into the Kindle Fire.... with whatever tweaks Amazon sees fit to add. I don't want to see WebOS die, and nobody has yet seen the Amazon version of Android, so it's unclear what we're going to be getting. I guess that will be the biggest question mark heading into November.... Amazon has no experience with non-web-based UI design. I'm sure they've got a ton of smart OS hackers making their cloud work, but that doesn't always translate into being able to create a nice UI. The "bookshelf" UI looks very attractive, but it might get clunky when you try to scale it. I'm not such a big fan of Apple's "Rows of Icons" UI, but it always runs smoothly and it's fairly easy to keep things organized and find what you want. I'm looking forward to AnandTech's review of the Kindle Fire.
  • JasonInofuentes - Wednesday, September 28, 2011 - link

    Remember that when Amazon caches a tablet optimized version of, say, Anandtech, it only updates the deltas throughout the day and can now offer that single cached copy to anyone that asks for it. So when millions of readers visit the site the cached copy is the perfect size to minimize bandwidth usage. And while you're right that this is a big commitment, Amazon has positions themselves to be a powerhouse in cloud services for some time. Their rates on storage and bandwidth alone are enough to draw clients as varied as Apple and Lilly.
    I will be curious if Silk shows up on other devices, phones for instance could take fuller advantage of the potential, especially when you consider that a decade after mobile versions of sites started to appear, most are anemic versions of their desktop counterparts.
    Thanks for the comment!
  • slashbinslashbash - Wednesday, September 28, 2011 - link

    Your example of Anandtech actually makes it worse for Amazon. I am logged in to AT and I am shown a personalized view of every page ("Welcome slashbinslashbash!"). There are a lot of sites that work like this, and for any of them the caching option just won't work -- Amazon will have no choice but to pull down a full version of each page the user loads. They can still cache graphics (which presumably don't change from user to user) but they are still going to require tons of storage and bandwidth to handle this. I know they're Amazon and their cloud is the biggest, but still, if they're going to ever make a profit off the Kindle Fire then they may want to reconsider whether Silk is a good solution. It is a small plus at best for the end-user, and a loss for Amazon.
  • batmanuel - Wednesday, September 28, 2011 - link

    That personalization is only a small element of the page. The graphics and article text are all going to be the same for various users, so those parts can be reused.
  • slashbinslashbash - Wednesday, September 28, 2011 - link

    The graphics are the same and won't be re-downloaded by Amazon, but the article text is part of the same HTML file that the personalized element is part of. Amazon can't send a request for just the top part of the HTML file from Anandtech's web servers, and Anandtech's servers can't send it; it's all or nothing (barring, of course, any AJAX components).

    In any case, it's still using bandwidth from Amazon to the end-user, which Amazon wouldn't have to pay for if they used a traditional browser on the Kindle. I know Amazon has bandwidth to spare, but it still costs them money in the end.
  • batmanuel - Wednesday, September 28, 2011 - link

    HTML is often the smallest part of most web pages. The HTML for this particular webpage is 123 kb. It's all of the other elements of the page: Flash elements, graphics, advertisements, stylesheets, etc. that can be reused, and those can be reused easily since they are always the same. Tiny HTML files are nothing for a company like Amazon that gives away online storage and music streaming.
  • batmanuel - Wednesday, September 28, 2011 - link

    That should read: ."..stylesheets, etc. that use up all the bandwidth."
  • empedocles - Wednesday, September 28, 2011 - link

    Actually, Silk is how Amazon will make money, a huge pile of money over time. Because of the way Silk works, Amazon will be able to capture complete web usage data for all the Kindle Fire users. They will then be able to correlate this with each person's content buying habits, content usage patterns, Amazon purchase info, Fire app usage info, etc. And Amazon will sell this highly personal data, similar to how Facebook sells highly detailed personal information on each user. Silk provides a high value revenue stream for Amazon and helps make it possible to offer the Kindle Fire at $199. Hence one can see that Silk is a critical component of the Fire's business model and that Fire might have been priced significantly higher if Amazon hadn't included Silk. Like it or not, selling user data seems to be par for the course for most tech companies. Silk is the technology that enables the capture of highly profitable web usage information for Amazon. To present it as anything else is at least in part disingenuous.
  • slashbinslashbash - Wednesday, September 28, 2011 - link

    Hmm, that has a depressing ring of truth to it. I wonder if it will be possible to disable that part of Silk. I for one am already tired enough of our Google overlords; I don't think I want to welcome Amazon as a secondary one. I wonder if they are going to make Silk available on other platforms too? You're right, that could end up being a significant revenue stream.

Log in

Don't have an account? Sign up now