1. Programming Massively Parallel Processors Lecture Slides for Chapter 9: Application Case Study – Electrostatic Potential Calculation
© David Kirk/NVIDIA and Wen-mei W. Hwu,
ECE408, University of Illinois, Urbana-Champaign

2. Motivation
Electrostatic potential map is used in building stable structures for molecular dynamics simulation.
© David Kirk/NVIDIA and Wen-mei W. Hwu,
ECE408, University of Illinois, Urbana-Champaign

3. Core Computation
The contribution of atom[i] to the electrostatic potential at lattice point j (potential[j]) is atom[i].charge / rij. In the Direct Coulomb Summation method, the total potential at lattice point j is the sum of contributions from all atoms in the system.
© David Kirk/NVIDIA and Wen-mei W. Hwu,
ECE408, University of Illinois, Urbana-Champaign

4. Sequential CPU Code © David Kirk/NVIDIA and Wen-mei W. Hwu, 2007-2010
ECE408, University of Illinois, Urbana-Champaign

5. DCS Kernel Design Overview
© David Kirk/NVIDIA and Wen-mei W. Hwu,
ECE408, University of Illinois, Urbana-Champaign

6. DCS Kernel Version 1 © David Kirk/NVIDIA and Wen-mei W. Hwu, 2007-2010
ECE408, University of Illinois, Urbana-Champaign

7. Information Reuse © David Kirk/NVIDIA and Wen-mei W. Hwu, 2007-2010
ECE408, University of Illinois, Urbana-Champaign

8. DCS kernel Version 2 © David Kirk/NVIDIA and Wen-mei W. Hwu, 2007-2010
ECE408, University of Illinois, Urbana-Champaign

9. Memory Layout for Coalescing
© David Kirk/NVIDIA and Wen-mei W. Hwu,
ECE408, University of Illinois, Urbana-Champaign

10. DCS Kernel Version 3 © David Kirk/NVIDIA and Wen-mei W. Hwu, 2007-2010
ECE408, University of Illinois, Urbana-Champaign

11. Performance Comparison
© David Kirk/NVIDIA and Wen-mei W. Hwu,
ECE408, University of Illinois, Urbana-Champaign

12. CPU vs. CPU-GPU Comparison
© David Kirk/NVIDIA and Wen-mei W. Hwu,
ECE408, University of Illinois, Urbana-Champaign

13. a PC with 3 GTX8800 GPUs at 700 Watts of total system power.
© David Kirk/NVIDIA and Wen-mei W. Hwu,
ECE408, University of Illinois, Urbana-Champaign