Wednesday 30/11/2011 Talks, Edinburgh Informatics Forum

A glance of blind computing

Vedran Dunjko

13:30, IF 4.02

Abstract: In 1978, Rivest et al. have, by asking “Is computation over data which has been encrypted possible?”, opened up a proliferate area of research in cryptography. The following 30 years yielded partial results in both the classical and new domain of quantum computation: Feigenbaum et al (1989) showed that classical computation with unconditional privacy of an NP-hard function is impossible (unless PH collapses to the 3rd level) and Childs (2005) and later Aharonov et. al, reflected on this problem in the quantum domain with only partial success. Then, 2009. saw breakthroughs in both settings: Gentry offered a positive answer in terms of a classical efficient fully homomorphic encryption, and Boradbent, Fitzsimmons and Kashefi presented the Universal Blind Quantum Computation (UBQC) protocol. Gentry’s classical scheme offers computational security whereas the UBQC scheme is unconditionally secure, but the user needs modest quantum powers. In this talk we will note the highlights of the turbulent history of computation with encrypted data, address the interplay between classical and quantum results, and their impact on cryptography, interactive proof systems and the understanding of the separation between “classical” and “quantum” in information processing. Finally, we will briefly go through the details of the UBQC protocol, and provide alternative proofs of its security.

The ZX-Calculus: a graphical approach to quantum computing

Ross Duncan

4pm, IF 4.31-33

Abstract: The ZX-calculus is a graphical notation for quantum computing based on monoidal categories and the physical notion of “strong complementarity”. In this talk I’ll explain what string complementarity is, and introduce the ZX-calculus. I’ll also demonstrate some recent applications of the calculus to problems in and around quantum computing.