A Graph Based Algorithm for Data Path Optimization in Custom Processors

Trajković, Jelena; Reshadi, Midia; Gorjiara, Bita; Gajski, Daniel D.

doi:10.1109/dsd.2006.7

Cited by 5 publications

(6 citation statements)

References 5 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Table 2 compares the total number of FPGA Logic Elements of the initial data path at the second column, which is produced by Spark Tool, and the data paths, which are the output of three optimizer algorithms. In order to compare our new refinement algorithm with other algorithms, we implemented Graph Based Algorithm (GBA) presented in [10] in our optimizer block in lieu of our refinement algorithm. To make a fair comparison, we provide identical conditions and constraints for all algorithms.…”

Section: Resultsmentioning

confidence: 99%

“…At first, the CDFG is used by the algorithm and is converted to a weighted directed acyclic graph (DAG).The conversion algorithm of CDFG to DAG is described in [10]. Then all paths from source nodes to sink node are extracted and sorted in a list with respect to their path lengths.…”

Section: Methodsmentioning

confidence: 99%

“…al. in [10] could successfully improve the optimization with the idea of separating the allocation and scheduling. But they could have reached a better optimization if they had taken other paths into account besides the critical path.…”

Section: Related Workmentioning

confidence: 98%

See 2 more Smart Citations

Data path refinement algorithm in high-level synthesis based on dynamic programming

Rahimi

Mohammadi

Foroughi

2009

2009 International Conference on Innovations in Information Technology (IIT)

View full text Add to dashboard Cite

As CMOS technology scales down into the deep-submicron domain, the cost of design, complexity and customization for Systems-On-Chip (SoCs) is rapidly increasing due to the inefficiency of traditional CAD tools. In this paper we present a new interactive refinement algorithm in high-level synthesis, based on dynamic programming, which maximizes resource optimization in data path. We start by quantifying the properties of the given application C code in terms of control data flow graph (CDFG), available parallelism and other metrics. We then apply designer guided constraints to a data path refinement algorithm for an initial data path. It attempts to reduce the number of the most expensive components while meeting the constraints. The experimental results show that not only the refined data path outperforms data paths refined by other heuristic methods, but also presents lower cost, less overhead and can be generated in less time.

show abstract

Section: Resultsmentioning

confidence: 99%

Section: Methodsmentioning

confidence: 99%

See 1 more Smart Citation

Data path refinement algorithm in high-level synthesis based on dynamic programming

Rahimi

Mohammadi

Foroughi

2009

2009 International Conference on Innovations in Information Technology (IIT)

View full text Add to dashboard Cite

show abstract

“…The derived required number of source operand transfers determines the number of register file output ports and the number of source busses, where the number of destination operand transfers translates into register file input ports and the destination buses [1]. The connection resources are allocated in greedy manner: output ports of all register files are connected to all source buses and input ports are connected to all destination busses.…”

Section: Introductionmentioning

confidence: 99%

Custom Processor Core Construction from C Code

Trajković

Gajski

2008

2008 Symposium on Application Specific Processors

View full text Add to dashboard Cite

In this paper we present a method for construction of application specific processor cores from a given C code. Our approach consists of three phases. We start by quantifying the properties of the C code in terms of operation types, available parallelism and other metrics. We then create an initial data path to exploit the available parallelism. We then apply designer guided constraints to an interactive data path refinement algorithm that attempts to reduce the number of the most expensive components while meeting the constraints. Our experimental results show that our technique scales very well with the size of the C code. We demonstrate the efficiency of our technique on wide range of applications, from standard academic benchmarks to industrial size examples like the MP3 decoder. Each processor core was constructed and refined in under a minute, allowing the designer to explore several different configurations in much less time than needed for manual design. On average, the refined core have only 23% latency overhead, twice as many block RAMs and 36% fewer slices compared to the respective manual designs.

show abstract

“…The heuristics use required number of source and destination operands and number of output and input ports for the storage elements available in the CL [Trajkovic et al, 2006].…”

Section: Initial Allocationmentioning

confidence: 99%

Automatic Data Path Generation from C code for Custom Processors

Trajković

Gajski

IFIP – The International Federation for Information Processing

View full text Add to dashboard Cite

Abstract:The stringent performance constraints and short time to market of modern digital systems require automatic methods for design of high performance applicationspecific architectures. This paper presents a novel algorithm for automatic generation of custom pipelined data path for a given application from its C code. The data path optimization targets both resource utilization and performance. The input to this architecture generator includes application C code, operation execution frequencies obtained by the profile run and a component library consisting of functional units, busses, multiplexers etc. The output is data path specified as a net-list of resource instances and their connections. The algorithm starts with an architecture that supports maximum parallelism for implementation of the input C code and iteratively refines it until an efficient resource utilization is obtained while maintaining the performance constraint. This paper also presents an algorithm to choose the priority of application basic blocks for optimization. Our experimental results show that automatically generated data paths satisfy given performance criteria and can be obtained in a matter of minutes leading to significant productivity gains.

show abstract

A Graph Based Algorithm for Data Path Optimization in Custom Processors

Cited by 5 publications

References 5 publications

Data path refinement algorithm in high-level synthesis based on dynamic programming

Data path refinement algorithm in high-level synthesis based on dynamic programming

Custom Processor Core Construction from C Code

Automatic Data Path Generation from C code for Custom Processors

Contact Info

Product

Resources

About