## A Kindle that will store every book ever printed

Every few years Amazon releases a new Kindle e-book reading device that includes more storage to hold more books. I personally have a Kindle 2 and am planning on picking up a 4 in the near future. I had the thought: Since the storage is exponentially increasing just like all information technology, how long until it could hold every book ever printed (let’s dream big). Google estimates that there are some 130, 000, 000 books that have been printed over the last few centuries, and since I’ll be dealing with non-illustrated figures provided by Amazon and the fact that Google might not have been totally thorough, I’ll set the bound as 200 000 000 books of pure text.

Now, Amazon lists how many non-illustrated books that each Kindle can store and ignoring the fact that they might be fudging a bit (since we’re just doing order of magnitude estimates for fun) I list them off thusly:

• Kindle 1 (2007) – 200 non-illustrated titles
• Kindle 2 (2009) – 1500 non-illustrated titles
• Kindle Touch (2011) – 3000 non-illustrated titles

I chose the Touch instead of the regular Kindle 4 since the 4 has reduced storage (only capable of holding 1400 titles) which I mark up to be an isolated non-ideality (given the low number of data points, perhaps unjustifiably). Also, had I included the slightly larger Kindle 3 (at 3500 titles), the projection would have been weird and I wanted to keep it simple.

The equation for the quadratic polynomial trendline is:

$\displaystyle 25(x-2006)^2 + 550(x-2006) - 375 = T$

Where x is the date in years and T is the total number of titles. Solving for $\displaystyle T = 200 000 000$ we arrive at the year 2817, and that’s assuming that no books are printed in the interim (not likely). That’s a fair ways off, but we are going from a very small data set and we might be in a “ramp-up” phase. 200 000 000 titles conservatively estimate at 1 MB/title gives a total information count of 200 TB stuffed into the Kindle form factor. Desktop hard drives are now at about 2 or 3 TB, so it’s understandable that cramming that amount of data into a tiny volume of solid state storage will take some time. But 800 years? Hmm…

It’s a totally academic exercise since the vast majority of those books would be uninteresting to any individual and you just wouldn’t have the time to read them all. Even assuming an outside case of a person who lives to be 100 and starts reading a book a day at the age of 5, they wouldn’t even crack 35 000 titles. Still, though, it would be nice to say “Hey human race. Remember the centuries of knowledge you accumulated and were able to cram the sum total into a space smaller than a dinner plate? I lost it at the airport!”

EDIT: Since it’s bothering me that I didn’t specifically enunciate it, I want to stress a big complication with this analysis is my use of a quadratic fitting function rather than a stronger exponential function, largely due to laziness. In a few years as more data points get added I hope to do a more comprehensive analysis. The main thrust of the argument is that don’t hold your breath for having every book ever written in your pocket (though you could cheat by using cloud storage I suppose).