In an article titled as above, a cluster was discussed that is able to brute-force guess any possible eight-character password containing upper- and lower-case letters, digits, and symbols, a total of 95^8 total possible combinations. It does this by guessing at a rate of 350 billion guesses a second. The cluster also uses other techniques, including searching a dictionary containing millions of words to optimize speed. This cluster was four times faster than the previous one due to an improvement in the algorithms used. The system cannot be used online due to the limits of guesses on accounts, but it may be used to brute-force conquer encryption methods that cannot be mathematically reversed but can cracked by running guesses through the same cryptographic function.
What I find interesting about this is the practical application of computing power and the scope of how fast computers today can tackle real-world problems. The generation of over 350 billion passwords seems almost inconceivably fast, but with computing power and efficient algorithms it is made possible. This article makes me think about how powerful computers will be in the future and how that power could possibly be applied. Big data immediately comes to mind, with the need for fast computing when a large amount of data needs to be analyzed or sorted through. Simulations that run using big data also raise a need for extremely fast computing. I decided to research a little to see what the needs are in this field, and I found an article on CNN.com about the likely emergence of a exaFLOP computer by the end of the decade. That means that a quintillion floating point operations could be done a second, equivalent to about 50 million laptops. A quote in the article states that “Aerospace engineering, astrophysics, biology, climate modeling and national security all have applications with extreme computing requirements”. Exascale could also “attempt to tackle very serious challenges in energy supply and sustainability”. I find it interesting that people can find applications that require such a scale of computing. The limit of computing power in the future seems like it will depend on the limitations of applications people can come up for it.
http://www.cnn.com/2012/03/29/tech/super-computer-exa-flop/index.html