Scalable FRM-SSA SoC Design for the Simulation of Networks with Thousands of Biochemical Reactions in Real Time

Abstract

Simulation of biomolecular networks with thousands of reactions is becoming essential for systems biology. We are presenting the design of a scalable System on Chip parallel architecture that implements Gillespie's First Reaction Method in reconfigurable FPGA hardware. Our SoC architecture can deliver performance (Mega-Reactions/sec) and throughput (MReaction cycles/sec) that is increasing linearly with the number of processors when simulating large biomolecular networks with up to m = 4096 reactions using a moderate size FPGA. We have synthesized and verified various SoC instances with up to N=8 Processing Elements for Xilinx Virtex 5 and Altera Cyclone III FPGAs, reaching clock frequencies up to 180 MHz and delivering simulation performance that is more than 2 order of magnitude higher than that of Intel Core 2 and i7 CPUs running at frequencies above 2GHz.

Python to accelerate embedded SoC design

“…In Hazapis and Manolakos [2011] we have shown how using parallel processing and pipelining it is possible to design a core that performs efficiently, stochastic simulations of biochemical reaction networks (BioModels) based on the First Reaction Method (FRM) of Gillespie's Stochastic Simulation Algorithm (SSA) [Gillespie 2007]. In this work we focus on how we can harness SysPy's design capabilities to build a flexible multi-processor SoC around the Leon3 processor utilizing an improved version of this SSA IP core component as an accelerator for systems biology simulations.…”

“…The IP core, which is captured in VHDL is parameterized by the number of PEs (N), the number of reactions in the BioModel (m) and other parameters and implements efficiently a parallelized version of Gillespie's FRM-SSA. Some recent improvements, compared to the original SSA core design presented in Hazapis and Manolakos [2011], are the following.…”

“…Very large networks are required to model the dynamical behavior of complete biological systems consisting of a large number of interlinked pathways. Modern FPGA devices with a large number of logic resources are good candidates to execute in parallel stochastic simulation of large-scale reaction networks [Hazapis and Manolakos 2011;Yoshimi et al 2006]. With the use of Python we managed to create a flexible platform, which scientists with a little or no experience in hardware design, can exploit to simulate their networks on an FPGA device attached to their PCs.…”

“…In Hazapis and Manolakos [2011] we have designed an IP core, with N processing elements (PEs) working in parallel, for simulating in reasonable amount of time largescale biochemical reacting networks. The IP core, which is captured in VHDL is parameterized by the number of PEs (N), the number of reactions in the BioModel (m) and other parameters and implements efficiently a parallelized version of Gillespie's FRM-SSA.…”

Scalable FPGA Accelerator of the NRM Algorithm for Efficient Stochastic Simulation of Large-Scale Biochemical Reaction Networks

“…Application-specific hardware architectures to accelerate the FRM SSA using FPGAs have been described in [4] and [5]. The implementation in [4] achieves a throughput of 10 Mega Reaction Cycles per sec (MRC/s) for small size biomodels.…”