## Iterative Methods for MPC on Graphical Processing Units

Publication: Research - peer-review › Conference abstract in proceedings – Annual report year: 2012

### Standard

**Iterative Methods for MPC on Graphical Processing Units.** / Gade-Nielsen, Nicolai Fog; Jørgensen, John Bagterp; Dammann, Bernd.

Publication: Research - peer-review › Conference abstract in proceedings – Annual report year: 2012

### Harvard

*Proceedings of the 17th Nordic Process Control Workshop.*Technical University of Denmark (DTU), Kogens Lyngby, pp. 161.

### APA

*Proceedings of the 17th Nordic Process Control Workshop*(pp. 161). Kogens Lyngby: Technical University of Denmark (DTU).

### CBE

### MLA

*Proceedings of the 17th Nordic Process Control Workshop.*Kogens Lyngby: Technical University of Denmark (DTU). 2012. 161.

### Vancouver

### Author

### Bibtex

}

### RIS

TY - ABST

T1 - Iterative Methods for MPC on Graphical Processing Units

AU - Gade-Nielsen,Nicolai Fog

AU - Jørgensen,John Bagterp

AU - Dammann,Bernd

PY - 2012

Y1 - 2012

N2 - The high oating point performance and memory bandwidth of Graphical Processing Units (GPUs) makes them ideal for a large number of computations which often arises in scientic computing, such as matrix operations. GPUs achieve this performance by utilizing massive par- allelism, which requires reevaluating existing algorithms with respect to this new architecture. This is of particular interest to large-scale constrained optimization problems with real-time requirements. The aim of this study is to investigate dierent methods for solving large-scale optimization problems with focus on their applicability for GPUs. We examine published techniques for iterative methods in interior points methods (IPMs) by applying them to simple test cases, such as a system of masses connected by springs. Iterative methods allows us deal with the ill-conditioning occurring in the later iterations of the IPM as well as to avoid the use of dense matrices, which may be too large for the limited memory capacity of current graphics cards.

AB - The high oating point performance and memory bandwidth of Graphical Processing Units (GPUs) makes them ideal for a large number of computations which often arises in scientic computing, such as matrix operations. GPUs achieve this performance by utilizing massive par- allelism, which requires reevaluating existing algorithms with respect to this new architecture. This is of particular interest to large-scale constrained optimization problems with real-time requirements. The aim of this study is to investigate dierent methods for solving large-scale optimization problems with focus on their applicability for GPUs. We examine published techniques for iterative methods in interior points methods (IPMs) by applying them to simple test cases, such as a system of masses connected by springs. Iterative methods allows us deal with the ill-conditioning occurring in the later iterations of the IPM as well as to avoid the use of dense matrices, which may be too large for the limited memory capacity of current graphics cards.

KW - Graphical Processing Unit

KW - Model based control

KW - Iterative methods

KW - Predictive control

KW - Optimization

M3 - Conference abstract in proceedings

SN - 978-87-643-0946-1

SP - 161

BT - Proceedings of the 17th Nordic Process Control Workshop

PB - Technical University of Denmark (DTU)

ER -