RHyMe: REDEFINE HyperCell Multicore for Accelerating HPC Kernels

In this paper, we present an architecture named REDEFINE HyperCell Multicore (RHyMe) designed to efficiently realize HPC application kernels, such as loops. RHyMe relies on the compiler to generate the meta-data for its functioning. Most of the orchestration activity for executing kernels is governed by compiler generated meta-data made use of at runtime. In RHyMe, macro operations can be realized as a hardware overlay of MIMO operations on hardware structures called HyperCells. While a HyperCell enables exploiting fine-grain instruction level and pipeline parallelism, coarse-grain parallelism is exploited among multiple HyperCells. Regularity exhibited by computations such as loops results in efficient usage of simple compute hardware such as HyperCells as well as memory structures that can be managed explicitly