11 citations found. Retrieving documents...
CHANG,J.AND PEDRAM, M. 1996. Module assignment for low power. In Proceedings of the EURODesign Automation Conference (Geneva, Switzerland). 376--381.

 Home/Search   Document Details and Download   Summary   Related Articles   Check  

This paper is cited in the following contexts:
Using Global Code Motions to Improve the Quality of .. - Gupta, Savoiu.. (2002)   (Correct)

....the number of resources means that the number of cliques cannot exceed the number of resources of each type. To solve this problem, we formulate it as a multi commodity network flow problem. A max cost flow through this multi commodity network represents a valid maximally 29 weighted clique cover [40, 41]. We note that Chang et al. 40] use the same formulation for module allocation but their objective is to minimize power consumption. 6.2 Variable to Register Binding Variable to register binding can take advantage of the improved operation binding by mapping variables that are inputs or outputs ....

....number of cliques cannot exceed the number of resources of each type. To solve this problem, we formulate it as a multi commodity network flow problem. A max cost flow through this multi commodity network represents a valid maximally 29 weighted clique cover [40, 41] We note that Chang et al. [40] use the same formulation for module allocation but their objective is to minimize power consumption. 6.2 Variable to Register Binding Variable to register binding can take advantage of the improved operation binding by mapping variables that are inputs or outputs to the same port of the same ....

J.-M. Chang and M. Pedram. Module assignment for low power. In European Design Automation Conference, 1996.


Dynamic Functional Unit Assignment for Low Power - Steve Haga Natasha   (Correct)

....processor ules. To begin, it is known [15] that the most important source of power dissipation in a module is the dynamic charging and discharging of its gates, called the switched capacitance. This switched capacitance is dependent upon the module s input values [14] It has been further shown [13, 6] that the Hamming distance of consecutive input patterns, defined as the number of bit positions that differ between them, provides a suitable measure of power consumption. In [13, 6] power is modeled as: # ( where ....

....capacitance. This switched capacitance is dependent upon the module s input values [14] It has been further shown [13, 6] that the Hamming distance of consecutive input patterns, defined as the number of bit positions that differ between them, provides a suitable measure of power consumption. In [13, 6], power is modeled as: # ( where = voltage, clock frequency, capacitance of output gate = average # transitions for output gate (called switching activity) the total capacitance ....

[Article contains additional citation context not shown here]

J.-M. Chang and M. Pedram. Module Assignment for Low Power. In Proc of the European Conference on Design Automation (EDAC), pages 376--381, September 1996.


Conditional Speculation and its Effects on.. - Gupta, Savoiu.. (2001)   (2 citations)  (Correct)

....of this graph will lead to binding that reduces interconnect. The constraint on the number of resources means that the number of cliques cannot exceed the number of resources of each type. To solve this problem, we formulate it as a multi commodity network flow problem and find the max cost flow [20, 21]. Chang et al. [20] use the same formulation for module allocation but their objective is to minimize power consumption. 6.2 Variable to Register Binding Variable to register binding can take advantage of the improved operation binding by mapping variables that are inputs or outputs to the same ....

....to binding that reduces interconnect. The constraint on the number of resources means that the number of cliques cannot exceed the number of resources of each type. To solve this problem, we formulate it as a multi commodity network flow problem and find the max cost flow [20, 21] Chang et al. [20] use the same formulation for module allocation but their objective is to minimize power consumption. 6.2 Variable to Register Binding Variable to register binding can take advantage of the improved operation binding by mapping variables that are inputs or outputs to the same port of the same ....

J.-M. Chang and M. Pedram. Module assignment for low power. In European Design Automation Conference, 1996.


Power Simulation and Estimation in VLSI Circuits - Pedram (1999)   (2 citations)  Self-citation (Pedram)   (Correct)

No context found.

J. M. Chang, M. Pedram, "Module Assignment for Low Power," EuroDAC-96: IEEE European Design Automation Conference, pp. 376-381, Geneva, Switzerland, September 1996.


High-Level Power Modeling, Estimation, and Optimization - Macii, Pedram, Somenzi (1997)   (22 citations)  Self-citation (Pedram)   (Correct)

No context found.

J. M. Chang, M. Pedram, "Module Assignment for Low Power," EuroDAC-96: IEEE European Design Automation Conference, pp. 376-381, Geneva, Switzerland, September 1996. 17


High-Level Power Modeling, Estimation, and Optimization - Macii, Pedram, Somenzi (1997)   (22 citations)  Self-citation (Pedram)   (Correct)

No context found.

J. M. Chang, M. Pedram, "Module Assignment for Low Power," IEEE EuroDAC-96, pp. 376-381, Geneva, Switzerland, Sep. 1996.


High-Level Power Modeling, Estimation, and Optimization - Macii, Pedram, Somenzi (1998)   (22 citations)  Self-citation (Pedram)   (Correct)

....the RT level techniques that will be described in Section IIC. Relevant data statistics such as the number of operations of a given type, bus and memory accesses, and I O operations are captured by static profiling based on stochastic analysis of the behavioral description and data streams [18] [19] or dynamic profiling based on direct simulation of the behavior under a typical input stream [20] 21] Instruction level or behavioral simulators are easily adapted to produce this information. C. RT Level Power Estimation Most RT level power estimation techniques use regressionbased, ....

....distributions, the way registers and functional units are allocated in the CDFG may heavily impact the switching activities at the interfaces of the resources. Graph based algorithms for register allocation for nonpipelined designs [64] and module allocation for functionally pipelined designs [19] proposed by Chang and Pedram rely on an accurate computation of the probability density functions at the inputs of the various resources, given the probability distributions for the system primary inputs. Unfortunately, in some cases, serial allocation may result in suboptimal solutions, i.e. ....

J. M. Chang and M. Pedram, "Module assignment for low power," in Proc. EuroDAC-96: IEEE Eur. Design Automation Conf., Geneva, Switzerland, Sept. 1996, pp. 376--381.


Optimal Simultaneous Module and - Multivoltage Assignment For   (Correct)

No context found.

CHANG,J.AND PEDRAM, M. 1996. Module assignment for low power. In Proceedings of the EURODesign Automation Conference (Geneva, Switzerland). 376--381.


Effective Techniques for the Generalized Low Power Binding.. - Davoodi, Srivastava   (Correct)

No context found.

J-M Chang, M. Pedram. "Module assignment for low power". In Proc. of European Design Automation Conf, pages 376--381, Sep. 1996.


Energy Estimation for Piecewise Regular Processor Arrays - Hannig, Teich (2002)   (Correct)

No context found.

Jui-Ming Chang and Massoud Pedram. Module Assignment for Low Power. In IEEE European Design Automation Conference (EuroDAC), pages 376--381, Geneva, Switzerland, September 1996.


Dynamic Functional Unit Assignment for Low Power - Haga, Reeves, Barua, Marculescu   (Correct)

No context found.

J.-M. Chang and M. Pedram. Module Assignment for Low Power. In Proc of the European Conference on Design Automation (EDAC), pages 376--381, September 1996.

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC