Unified Parallel C (UPC) Comes To LLVM/Clang

Clang UPC has been announced, which is a Unified Parallel C implementation targeting the LLVM/Clang compiler stack. Unified Parallel C is a C99 extension targeting high-performance computing on parallel machines.

Unified Parallel C is coming to GCC after its support was living separately from mainline GCC for some time. Now, developers have announced an initial implementation of Unified Parallel C for LLVM's Clang compiler. Clang UPC is this initial implementation.

Aside from needing the separate Clang UPC code-base, for now a special version of LLVM is required. This branched version of LLVM has a few minor changes necessary for supporting Unified Parallel C that aren't yet in the mainline code-base.

UPC is described as "an extension of the C programming language designed for high-performance computing on large-scale parallel machines, including those with a common global address space (SMP and NUMA) and those with distributed memory (e.g. clusters). The programmer is presented with a single shared, partitioned address space, where variables may be directly read and written by any processor, but each variable is physically associated with a single processor. UPC uses a Single Program Multiple Data (SPMD) model of computation in which the amount of parallelism is fixed at program startup time, typically with a single thread of execution per processor."

Eventually there is a plan to merge Clang UPC into the main Clang/LLVM code-base. In addition, there is a desire to develop a UPC-to-C source translator based upon Clang UPC. This work though is too late for the LLVM/Clang 3.2 release that will happen in December.

For more details on the Clang UPC announcement, see the Clang mailing list.