"TREBUCHET: A High-Powered Processor for Cutting-Edge Encryption"

Fully Homomorphic Encryption (FHE) enables algorithms to do direct computations on encrypted data. Usually, sensitive data is encrypted, and it must be decrypted before it can be used for any form of analysis or computing. The analysis or computation is conducted while the sensitive data is in an unencrypted state, and then the data is re-encrypted. Matthew French, Research Director at USC Viterbi's Information Sciences Institute (ISI), says that the problem with these schemes is that there is inevitably a breakdown in the process, and someone can snoop on the unencrypted processing, or someone can forget to re-encrypt the data. In the past decade, breakthrough advances in algorithms have enabled FHE, which eliminates the need to decrypt and re-encrypt data, resulting in a far more secure system, according to French. However, FHE requires substantially more computational power to accomplish tasks equivalent to those that are not encrypted. FHE requires around 100,000 times more processing than conventional techniques, so FHE must decrease the computation gap in order to be useful. French and his colleagues took on the challenge with their co-processor, TREBUCHET, which addresses this by developing custom computer hardware to accelerate FHE processing with the aim of achieving ten times the speed of traditional processing. TREBUCHET was created for the Data Protection in Virtual Environments (DPRIVE) Program of the Defense Advanced Research Projects Agency (DARPA). There are both private research facilities and academic institutions on the team. This article continues to dicuss the concept of FHE and the TREBUCHET solution. 

USC Viterbi reports "TREBUCHET: A High-Powered Processor for Cutting-Edge Encryption"