Dataflow Computing with Polymorphic Registers
Paper i proceeding, 2013

Heterogeneous systems are becoming increasingly popular for data processing. They improve performance of simple kernels applied to large amounts of data. However, sequential data loads may have negative impact. Data parallel solutions such as Polymorphic Register Files (PRFs) can potentially accelerate applications by facilitating high speed, parallel access to performance-critical data. Furthermore, by PRF customization, specific data path features are exposed to the programmer in a very convenient way. PRFs allow additional control over the registers dimensions, and the number of elements which can be simultaneously accessed by computational units. This paper shows how PRFs can be integrated in dataflow computational platforms. In particular, starting from an annotated source code, we present a compiler-based methodology that automatically generates the customized PRFs and the enhanced computational kernels that efficiently exploit them.

dataflow computing

Field programmable gate arrays

polymorphic register files

compiler-based methodology

Författare

Catalin Ciobanu

Chalmers, Data- och informationsteknik, Datorteknik

Georgi Gaydadjiev

Chalmers, Data- och informationsteknik, Datorteknik

Christian Pilato

Politecnico di Milano

Donatella Sciuto

Politecnico di Milano

International Conference on Embedded Computer Systems: Architectures, Modeling, and Simulation (SAMOS XIII), 2013

314 - 321 6621140

Facilitating Analysis and Synthesis Technologies\nfor Effective Reconfiguration (FASTER)

Europeiska kommissionen (FP7), 2011-09-01 -- 2014-11-30.

Ämneskategorier

Datorteknik

Inbäddad systemteknik

Datorsystem

Styrkeområden

Informations- och kommunikationsteknik

DOI

10.1109/SAMOS.2013.6621140

ISBN

978-1-4799-0103-6

Mer information

Senast uppdaterat

2018-03-29