On the Limitations of Power Macromodeling Techniques

Klein, Felipe; Araújo, Guido; Azevedo, Rodolfo; Leao, R.; Santos, Luiz C. V. dos

doi:10.1109/isvlsi.2007.74

Cited by 4 publications

(3 citation statements)

References 14 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This methodology tends to give large errors when two different input signals that result in different output statistics, are characterized with the same signal parameters [8].…”

Section: Related Workmentioning

confidence: 99%

“…Consider the following three vectors: 010010110 010011010 010110100 (8) We say that the first two belong to the Gaussian distribution with the most significant mean bits equal to 01001, and the second and the third belong to the Gaussian distribution with the most significant mean bits equal to 010. In both cases, the first bit that stands immediately after the mean bits, changes with a switching activity of 1, and the rest of the bits behave as uncorrelated, random bits with a switching activity of 0.5.…”

Section: Cycle-by-cycle Accuracymentioning

confidence: 99%

See 1 more Smart Citation

A Comparison of Approaches for High-Level Power Estimation of LUT-Based DSP Components

Jevtić

Carreras

Helms

2008

2008 International Conference on Reconfigurable Computing and FPGAs

View full text Add to dashboard Cite

show abstract

“…This methodology tends to give large errors when two different input signals that result in different output statistics, are characterized with the same signal parameters [8].…”

Section: Related Workmentioning

confidence: 99%

Section: Cycle-by-cycle Accuracymentioning

confidence: 99%

A Comparison of Approaches for High-Level Power Estimation of LUT-Based DSP Components

Jevtić

Carreras

Helms

2008

2008 International Conference on Reconfigurable Computing and FPGAs

View full text Add to dashboard Cite

show abstract

“…Architectural power simulators are typically used in early design phases when circuit details are not yet known. Complex circuits consisting of thousands of gates are often abstracted into simple power macro models [2], [3]. Abstracting away individuals signals provides a tremendous reduction of simulation time and allows architectural power simulators to estimate power consumption for workloads that are out of the reach of circuit or RTL power simulators.…”

Section: Introductionmentioning

confidence: 99%

ALUPower: Data Dependent Power Consumption in GPUs

Lucas

Juurlink

2016

2016 IEEE 24th International Symposium on Modeling, Analysis and Simulation of Computer and Telecommunication Systems (MASCOTS)

View full text Add to dashboard Cite

Abstract-Existing architectural power models for GPUs count activities such as executing floating point or integer instructions, but do not consider the data values processed. While data value dependent power consumption can often be neglected when performing architectural simulations of high performance Out-of-Order (OoO) CPUs, we show that this approach is invalid for estimating the power consumption of GPUs. The throughput processing approach of GPUs reduces the amount of control logic and shifts the area and power budget towards functional units and register files. This makes accurate estimations of the power consumption of functional units even more crucial than in OoO CPUs. Using measurements from actual GPUs, we show that the processed data values influence the energy consumption of GPUs significantly. For example, the power consumption of one kernel varies between 155 and 257 Watt depending on the processed values. Existing architectural simulators are not able to model the influence of the data values on power consumption. RTL and gate level simulators usually consider data values in their power estimates but require detailed modeling of the employed units and are extremely slow. We first describe how the power consumption of GPU functional units can be measured and characterized using microbenchmarks. Then measurement results are presented and several opportunities for energy reduction by software developers or compilers are described. Finally, we demonstrate a simple and fast power macro model to estimate the power consumption of functional units and provide a significant improvement in accuracy compared to previously used constant energy per instruction models.

show abstract