Fanout decomposition dataflow optimizations for FPGA-based Sparse LU factorization
Performance of FPGA-based token dataflow architectures is often limited by the long tail distribution of parallelism in the compute paths of the dataflow graphs. This is known to limit speedup of dataflow processing of Sparse LU factorization to only 3-10x over CPUs. One reason behind the limitation...
Main Authors: | Siddhartha, Kapre, Nachiket |
---|---|
Other Authors: | School of Computer Engineering |
Format: | Conference Paper |
Language: | English |
Published: |
2015
|
Subjects: | |
Online Access: | https://hdl.handle.net/10356/81207 http://hdl.handle.net/10220/39179 |
Similar Items
-
Heterogeneous dataflow architectures for FPGA-based sparse LU factorization
by: Siddhartha, et al.
Published: (2015) -
Breaking Sequential Dependencies in FPGA-Based Sparse LU Factorization
by: Siddhartha, et al.
Published: (2015) -
Limits of Statically-Scheduled Token Dataflow Processing
by: Kapre, Nachiket, et al.
Published: (2015) -
Custom FPGA-based soft-processors for sparse graph acceleration
by: Kapre, Nachiket
Published: (2015) -
An NoC Traffic Compiler for Efficient FPGA Implementation of Sparse Graph-Oriented Workloads
by: Kapre, Nachiket, et al.
Published: (2015)