Hardware/software partitioning using integer programming

Abstract

One of the key problems in hardware software c odesign is hardware software p artitioning. This paper describes a new approach to hardware software p artitioning using integer programming IP. The advantage of using IP is that optimal results are c alculated r espective to the chosen objective function. The partitioning approach works fully automatic and supports multi-processor systems, interfacing and hardware sharing. In contrast to other approaches where special estimators are used, we use compilation and synthesis tools for cost estimation. The increased time for calculating the cost metrics is compensated by an improved quality of the estimations compared to the results of estimators. Therefore, fewer iteration steps of partitioning will be n e eded. The paper will show that using integer programming to solve the hardware software p artitioning problem is feasible and leads to promising results.

Knapsack Model and Algorithm for HW/SW Partitioning Problem

“…A software oriented approach has been proposed by Ernst et al [3], Vahid et al [4]. Hardware oriented approach has been proposed in Gupta et al [5], Niemann et al [6]. In [9], the authors proposed a flexible granularity approach for hardware software partitioning.…”

Evaluating the Kernighan-Lin Heuristic for Hardware/Software Partitioning

“…In particular, many researchers consider scheduling as part of partitioning (Chatha and Vemuri, 2001;Dick and Jha, 1998;Kalavade and Lee, 1997;Lopez-Vallejo and Lopez, 2003;Mei et al, 2000;Niemann and Marwedel, 1997), whereas others do not (Eles et al, 1996;Grode et al, 1998;Madsen et al, 1997;O'Nils et al, 1995;Vahid and Le, 1997;Vahid, 2002). Some even include the problem of assigning communication events to links between hardware and/or software units (Dick and Jha, 1998;Mei et al, 2000).…”

“…The proposed exact algorithms include branch-and-bound (Binh et al, 1996), dynamic programming (Madsen et al, 1997;O'Nils et al, 1995), and integer linear programming (Mann and Orbán, 2003;Niemann, 1998;Niemann and Marwedel, 1997).…”

Hardware/software partitioning with iterative improvement heuristics

“…In [6,8,13,16] automatic partitioning is performed, while the approach presented in [1] is based on manual partitioning. Partitioning at a fine grained level is performed in [6,8].…”