Distributed.net cracked the RC5-64 encryption algorithm after four years, proving it's unsuitable for long-term security. The project demonstrates the massive potential of distributed computing for solving computationally expensive cryptographic challenges.
distributed.net cracked RC5-64. RC5-64 is an encryption algorithm developed by RSA. It took 4 years. It looks long. But the big deal is, the algorithm therefore can’t be used for long-term security. More importantly, it’s becoming practical to use distributed computing on a massive scale.
I observed IBM’s move into on-demand computing under Sam Palmisano, validating my prediction that tech giants would soon aggregate retail computing power for massive distributed projects, similar to early efforts by Google and Intel.
I examine the rise of commercial distributed computing markets where companies buy idle retail processor time. This model dis-aggregates computing power, enabling resource-intensive initiatives like the Internet Movie Project to be crowdsourced using spare capacity.
CERN is adopting distributed computing models, similar to SETI@home, to process the massive datasets generated by particle physics experiments. This transition leverages global idle processing power to handle the computational demands of the Large Hadron Collider.
Distributed Proofreaders uses a web-based crowdsourcing model to digitize books for Project Gutenberg. By leveraging volunteer labor, the project proofreads over 1,000 pages daily to preserve and distribute public domain texts more efficiently.
I explored the hierarchy of the PC industry through a look at 64-bit processors. Seeing Microsoft atop Intel, HP, and Dell provides a useful framework for predicting how the market and hardware standards will evolve over time.