A Modified Merging Approach for Datapath Configuration Time Reduction

This paper represents a modified datapath merging technique to amortize the configuration latency of mapping datapaths on reconfigurable fabric in Run-Time Reconfigurable Systems (RTR). This method embeds together the different Data Flow Graphs (DFGs), corresponding to the loop kernels to create a single datapath (merged datapath) instead of multiple datapaths. The DFGs are merged in steps where each step corresponds to combining a DFG onto the merged datapath. Afterwards, the method combines the resources inside the merged datapath to minimize the configuration time by employing the maximum weighted clique technique. The proposed merging technique is evaluated using the Media-bench suit workloads. The results indicate that our technique outperforms previous HLS approaches aimed at RTR systems and reduces the datapath configuration time up to 10%.

[1]  Zhiyuan Li,et al.  Configuration management techniques for reconfigurable computing , 2002 .

[2]  Katherine Compton,et al.  An execution environment for reconfigurable computing , 2005, 13th Annual IEEE Symposium on Field-Programmable Custom Computing Machines (FCCM'05).

[3]  Mahmood Fazlali,et al.  Data path Configuration Time Reduction for Run-time Reconfigurable Systems , 2009, ERSA.

[4]  Philippe Coussy,et al.  High-Level Synthesis: from Algorithm to Digital Circuit , 2008 .

[5]  Stamatis Vassiliadis,et al.  DWARV: Delftworkbench Automated Reconfigurable VHDL Generator , 2007, 2007 International Conference on Field Programmable Logic and Applications.

[6]  Mahmood Fazlali,et al.  A New Datapath Merging Method for Reconfigurable System , 2009, ARC.

[7]  Miodrag Potkonjak,et al.  MediaBench: a tool for evaluating and synthesizing multimedia and communications systems , 1997, Proceedings of 30th Annual International Symposium on Microarchitecture.

[8]  Cid C. de Souza,et al.  Efficient datapath merging for partially reconfigurable architectures , 2005, IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems.

[9]  Patric R. J. Östergård,et al.  A New Algorithm for the Maximum-Weight Clique Problem , 1999, Electron. Notes Discret. Math..

[10]  Kaushik Roy,et al.  Synthesis of application-specific highly efficient multi-mode cores for embedded systems , 2005, TECS.

[11]  Mahmood Fathy,et al.  A new compression based approach for reconfiguration overhead reduction in virtex based RTR systems , 2006, Comput. Electr. Eng..

[12]  Stamatis Vassiliadis,et al.  A Quantitative Prediction Model for Hardware/Software Partitioning , 2007, 2007 International Conference on Field Programmable Logic and Applications.

[13]  Jari Nurmi,et al.  Design Flow Instantiation for Run-Time Reconfigurable Systems: A Case Study , 2008, EURASIP J. Embed. Syst..

[14]  Majid Sarrafzadeh,et al.  An optimal algorithm for minimizing run-time reconfiguration delay , 2004, TECS.

[15]  David S. Johnson,et al.  Computers and Intractability: A Guide to the Theory of NP-Completeness , 1978 .

[16]  Philippe Coussy,et al.  High-Level Synthesis , 2008 .

[17]  Stamatis Vassiliadis,et al.  The MOLEN polymorphic processor , 2004, IEEE Transactions on Computers.

[18]  Markus Rullmann,et al.  A cost model for partial dynamic reconfiguration , 2008, 2008 International Conference on Embedded Computer Systems: Architectures, Modeling, and Simulation.