PetaBricks is a new implicitly parallel language and compiler where having multiple implementations of multiple algorithms to solve a problem is the natural way of programming. We make algorithmic choice a fundamental construct of the language. Choices are provided in a way that also allows our compiler to tune at a finer granularity. The PetaBricks compiler autotunes programs by making both fine-grained as well as algorithmic choices. Choices also include different automatic parallelization techniques, data distributions, algorithmic parameters, transformations, and blocking.
In addition, we introduce novel techniques to autotune algorithms for different convergence criteria. When choosing between various direct and iterative methods, the PetaBricks compiler is able to tune a program in such a way that delivers near-optimal efficiency for any desired level of accuracy. The compiler has the flexibility of utilizing different convergence criteria for the various components within a single algorithm, providing the user with accuracy choice alongside algorithmic choice.
If you have questions you can contact Jason Ansel.
PetaBricks is open source and is available under the MIT license. To try PetaBricks yourself, see our getting started page.