Twitter distributes software 75 times faster using BitTorrent
My business associate developed the BitTornado software Twitter is using to solve their server update problem. It’s the same type of problem Apple had when I worked there in 1992. The solution approach similar.
A single distribution server is a bottleneck both for efficiency and reliability. A single server slows down as more clients are accessing it. If the single server fails the entire process stops.
Predictive Innovation defines 15 Alternatives for any problem. The 15 Alternatives are arranged into two dimensions of 3 scales and 5 directions. The 3 scales are Single, Multiple and Continuous (see “Predicting the Next Disruptive Innovation” for more details).
When I was at Apple we solved the problem by moving from a Single mainframe to multiple servers and clients. This spread the bandwidth across more computers and provided more flexibility for doing ad-hoc reports.
BitTorrent is a Continuous solution. It doesn’t just spread the bandwidth over more computers, it spreads it over all the computers using the torrent. It produces the maximum bandwidth with the least vulnerability. It also flips the concept of speed on its head. Instead of getting slower as more computers download a file, it gets faster. Any computer connected to the torrent can share pieces with any other computer connected to the torrent. This is also how it reduces vulnerability. There are multiple sources for the pieces of the file, even if none of them have a complete copy of the file.
Electric Drag Racer Sets Records
Plugin electric car beats gasoline muscle cars at drag racing. White Zombie set records with simple lead acid batteries.
The fact this car was setting speed records using simple, reliable, cheap, lead acid batteries is the major innovation. The definition of innovation is:
Profitably satisfy unmet desires
Profitably doesn’t necessarily mean money, it means all the desires are better satisfied for everyone involved in delivering the innovation. So far plugin electric cars fail to innovate as normal passenger cars because the range is less than gasoline, diesel, or natural gas cars. Driving very fast for short distances is a way for plugin electrics to innovate. Electric cars have more torque from the start than internal combustion engines, this makes them very fast off the line.
Finding a special use for a new technology is part of the evolution to a new paradigm. In the beginning internal combustion engines had severe limitations but eventually outperformed the previous technology in every way and even made new things possible. Airplanes became possible because of internal combustion engines.
Electric vehicles have a large amount of torque, are quiet, and don’t any toxic exhaust fumes while operating. This makes electric vehicles ideal for use indoors especially hi-lo to move heavy items in a warehouse. Electric vehicles work well indoors because they don’t need to travel far and can be recharged in between uses. This has been how plugin electric vehicles have been used for decades. The other use is for golf carts. Again, a golf cart has limited range and is quiet which is perfect for a golf course.
Mapping the steps from current technology and uses to the replacement and new possibilities reveals the lowest risk, highest profit innovations.