PipeRench: a coprocessor for streaming multimedia acceleration

Abstract

Future computing workloads will emphasize an architecture's ability to perform relatively simple calculations on massive quantities of mixed-width data. This paper describes a novel reconfigurable fabric architecture, PipeRench, optimized to accelerate these types of computations. PipeRench enables fast, robust compilers, supports forward compatibility, and virtualizes configurations, thus removing the fixed size constraint present in other fabrics. For the first time we explore how the bit-width of processing elements affects performance and show how the PipeRench architecture has been optimized to balance the needs of the compiler against the realities of silicon. Finally, we demonstrate extreme performance speedup on certain computing kernels (up to 190x versus a modern RISC processor), and analyze how this acceleration translates to application speedup.

A reconfigurable unit for a clustered programmable-reconfigurable processor

“…It also leads to a counter-clockwise flow of data through the reconfigurable cluster, similar to the structure found in PipeRench [5]. Each register has a write enable input driven by the array control unit (ACU).…”

“…The datapath is formed by a row of coarse-grain word-based functional units connected to a programmable interconnect with a variable configurable length. PipeRench [5] is another coarse-grain architecture that organizes its functional units into multiple rows, or stripes, consisting of 8 bit ALUs, registers and a local communication interface. Similar to Amalgam's RC, global communication is performed by a vertical global bus which gives the architecture a direction of computational flow.…”

A Survey on Dynamically Reconfigurable Processors

“…Then, the streaming data is inserted and executed in the pipelined manner. Kilocore KC256 [8], which is a commercial version of PipeRench [17] has specialized structure for the pipelined execution. As shown in Fig.…”

Speedups and Energy Savings of Microprocessor Platforms with a Coarse-Grained Reconfigurable Data-Path

“…Embedded systems have been proposed that extend the computational capabilities of a microprocessor core by coupling it with reconfigurable logic [1]- [8]. Reconfigurable hardware is used to implement computational intensive parts of the applications, called kernels.…”