The Rhind papyrus is probably the oldest written document that we have, it goes back over 3650 years. This is an amazing lifetime of information storage. Nothing we have today can compare to that. Our CDs are good for a few decades, mag tapes if properly preserved can last a bit longer but actually there is no assurance that any digital media can last over 100 years.
Physics last year proposed a new way for coding information optically and by experiment they came up to a staggering stability of information under stressful condition of 13.8 billion years, the age of the universe. The catch is that so far the method is not practical for its slow writing speed. In general, it is also true that there is a tradeoff between the density of information stored and their lifetime. As an example scientists have been able to store bits on a single atom, and that would result in an amazing density (at least for orders of magnitude compared to today high density storage), but the atoms can retain the information for about one billionth of second.
Now a team of researchers at ETH Zurich has demonstrated a way to store information by coding it on a DNA strand. This is not the first time researchers are using DNA to store information (and Nature chose DNA long ago exactly for this purpose) but it is the first time that a process has been created for an industrial adoption. The DNA is very resilient in terms of preserving information, we can read the DNA of long lost creatures going back million of years. This is because of its intrinsically strong binds and because of the duplication of information making sure that if something degrades it can be spotted and a good copy can be found at its place.
The researchers at ETH have demonstrated the feasibility by storing the Switzerland Federal Charter and the Methods of Mechanical Theorems (of Archimede) onto a strand of DNA that has been encapsulated in a silica ball with a diameter of about 150nm. That means several orders of magnitude denser than today's flash memories and several orders of magnitude in life time.
I am an admirer of Nature and of its capability to get smarter and smarter of time through natural selection processes. Seeing scientists learning from Nature is a real pleasure.