1.  Automatic Parallelization of Programs

2. A taxonomy of pairing-friendly elliptic curves

1.  Automatic Parallelization of Programs

In this project, we propose to build on our previous work on the transformation of functional programs to automatically parallelize the programs resulting from these transformations. Our program transformation algorithm, which we call distillation, produces a normalised form of program which is very amenable to parallel execution. We therefore propose to further transform the normalised program into a parallel equivalent. In order to evaluate the efficiency of the resulting parallel programs in comparison to other parallelization techniques, we will need to run them in a parallel environment such as that which would be provided by the cluster.

2. A taxonomy of pairing-friendly elliptic curves

Dr. Scott had a class B project with ICHEC (finished on 18th October), used in the hunt to find pairing friendly elliptic curves. We now want to renew the search, as we have some new ideas. In fact we did not find any new curves in our ICHEC search. However smaller scale searches (overnight runs on my own PC) did find some new ones, as reported below*. This paper is under submission to the Journal of Cryptography. The fact that the ICHEC search did not find any new curves strengthens our view that in fact our "taxonomy" as reported in the paper is fairly complete.