CODES+ISSS '06 Proceedings of the 4th international conference on Hardware/software codesign and system synthesis
SoCDAL: System-on-chip design AcceLerator
ACM Transactions on Design Automation of Electronic Systems (TODAES)
Validating power architecture™ technology-based MPSoCs through executable specifications
IEEE Transactions on Very Large Scale Integration (VLSI) Systems
Trace-driven optimization of networks-on-chip configurations
Proceedings of the 47th Design Automation Conference
Simulation-based verification of the MOST NetInterface specification revision 3.0
Proceedings of the Conference on Design, Automation and Test in Europe
Combined system synthesis and communication architecture exploration for MPSoCs
Proceedings of the Conference on Design, Automation and Test in Europe
Journal of Systems Architecture: the EUROMICRO Journal
Hi-index | 0.00 |
As billion transistor System-on-chips (SoC) become commonplace and design complexity continues to increase, designers are faced with the daunting task of meeting escalating design requirements in shrinking time-to-market windows, and have begun using an IP-based SoC design methodology that permits reuse of key SoC functional components. Since the communication architectures connecting components in these SoC designs significantly impact system performance, it is imperative that designers explore the communication design space efficiently, quickly and early in the design flow. Transaction Level Modeling (TLM) is an emerging abstraction that facilitates early exploration of SoC architectures. This paper outlines a typical IP-based SoC design flow, and presents the Cycle Count Accurate at Transaction Boundaries (CCATB) modeling abstraction which is a fast, efficient and flexible approach for exploring bus-based communication architectures in SoC designs. The CCATB models not only take less time to model but are also faster to simulate than existing modeling abstractions for communication architecture exploration such as pin-accurate BCA (PABCA) and transaction based BCA (T-BCA). Experimental results on several industrial SoC subsystem case studies show that CCATB models are faster than PA-BCA by as much as 120% on average and by 67% on average when compared to T-BCA, demonstrating the advantages of CCATB-based TLM abstraction for exploring bus-based SoC communication architectures.