A Polymorphic Register File for matrix operations

Abstract

Abstract-Previous vector architectures divided the available register file space in a fixed number of registers of equal sizes and shapes. We propose a register file organization which allows dynamic creation of a variable number of multidimensional registers of arbitrary sizes referred to as a Polymorphic Register File. Our objective is to evaluate the performance benefits of the proposed organization. Simulation results using real applications (Floyd and CG) suggest speedups of up to 3 times compared to the Cell SPU for Floyd and 2 times compared to a one dimensional vectorized version of the sparse matrix vector multiplication. Moreover, in the same experimental context, a large reduction in the number of executed instructions of up to 3000 times for Floyd and 2000 times for sparse matrix vector multiplication is achieved.

Scalability Study of Polymorphic Register Files

“…Currently, only 1D and 2D operands are supported, but the PRF can be extended for any number of dimensions. Potential performance gains are due to multiaxis vectorization, efficient register storage utilization, higher Previous works indicate possible reductions of the number of executed instructions by three orders of magnitude due to PRF [4]. Furthermore, PRFs allow performance benefits when compared to the Cell processor for Floyd and the main kernel of the CG Method -sparse matrix vector multiplication [4].…”

“…Potential performance gains are due to multiaxis vectorization, efficient register storage utilization, higher Previous works indicate possible reductions of the number of executed instructions by three orders of magnitude due to PRF [4]. Furthermore, PRFs allow performance benefits when compared to the Cell processor for Floyd and the main kernel of the CG Method -sparse matrix vector multiplication [4]. The PRF programming interface allows high performance dense matrix multiplication with at least 35 times less instructions than a hand-crafted version for the Cell BE [14].…”

“…The PRF registers are multidimensional, with arbitrary sizes and can be created / resized at runtime. Previous studies ( [4], [14]) have demonstrated that PRFs suit computationally intensive workloads such as Floyd, the Conjugate Gradient (CG) method and dense matrix multiplication. Moreover, PRFs could improve performance and efficiency in state of the art many-core computers, potentially saving area and power as shown in [5].…”

“…The Polymorphic Register File (PRF) [4] was proposed to provide a relaxed view for the programmer of high performance vector applications. Unlike in existing vector architectures, such as IBM 370 [3] and the Synergistic Processor Units in the Cell BE [9], the PRF storage is not divided in a fixed number of equally sized registers.…”

“…BACKGROUND AND RELATED WORK A PRF is a parameterizable register file, logically reorganized under software control, by the system / application programmer or by the runtime system, to support multiple register dimensions and sizes simultaneously [4]. Fig.…”