Improving the energy efficiency of sparse linear system solvers on multicore and manycore systems

  title={Improving the energy efficiency of sparse linear system solvers on multicore and manycore systems},
  author={Hartwig Anzt and Enrique S. Quintana‐Ort{\'i}},
  journal={Philosophical Transactions of the Royal Society A: Mathematical, Physical and Engineering Sciences},
  • H. AnztE. S. Quintana‐Ortí
  • Published 28 June 2014
  • Computer Science
  • Philosophical Transactions of the Royal Society A: Mathematical, Physical and Engineering Sciences
While most recent breakthroughs in scientific research rely on complex simulations carried out in large-scale supercomputers, the power draft and energy spent for this purpose is increasingly becoming a limiting factor to this trend. In this paper, we provide an overview of the current status in energy-efficient scientific computing by reviewing different technologies used to monitor power draft as well as power- and energy-saving mechanisms available in commodity hardware. For the particular… 

Figures and Tables from this paper

Experiences in autotuning matrix multiplication for energy minimization on GPUs

This paper reports extensive results and analysis of autotuning the computationally intensive graphics processing units kernel for dense matrix–matrix multiplication in double precision, and takes the energy efficiency into account.

The ICARUS White Paper: A Scalable, Energy-Efficient, Solar-Powered HPC Center Based on Low Power GPUs

By this work, it is shown, that with current technology, energy consumption-induced follow-up cost of HPC can be reduced to zero.

Adaptive precision solvers for sparse linear systems

An implementation of a Jacobi iterative solver for sparse linear systems that iterates the distinct components of the solution with different precision in terms of mantissa length is formulated and a novel metric is introduced which quantifies the average mantissalength during the iteration.

Energy efficiency of the simulation of three-dimensional coastal ocean circulation on modern commodity and mobile processors

We analyze energy efficiency of a 3D coastal ocean simulator on Haswell and Cortex-A15 architectures and propose a simple yet effective way to model energy-to-solution on different hardware

How applied sciences can accelerate the energy revolution

It is demonstrated how the applied sciences community can make a significant contribution in reducing the energy footprint of their computations.



Performance and Energy Analysis of the Iterative Solution of Sparse Linear Systems on Multicore and Manycore Architectures

This analysis employs the conjugate gradient method, an important algorithm for the iterative solution of linear systems that is basically composed of the sparse matrix-vector product and other (minor) vector kernels.

Reformulated Conjugate Gradient for the Energy-Aware Solution of Linear Systems on GPUs

A redesign of the conjugate gradient method for the iterative solution of sparse linear systems on heterogeneous systems accelerated by graphics processing units (GPUs) results in a slight increase of performance and enables the efficient exploitation of power-saving techniques implicit in the hardware.

Optimization of power consumption in the iterative solution of sparse linear systems on graphics processors

Analyzing the power consumption of different GPU-accelerated iterative solver implementations enhanced with energy-saving techniques reveals how the usage of idle-wait during GPU-kernel calls effectively leverages the power-tools provided by hardware, and improves the energy performance of the algorithm.

PowerPack: Energy Profiling and Analysis of High-Performance Systems and Applications

A framework to isolate the power consumption of devices including disks, memory, NICs, and processors in a high-performance cluster and correlate these measurements to application functions is extended and conclusively how intelligent DVFS scheduling can enhance system energy efficiency while maintaining performance is revealed.

Tools for Power-Energy Modelling and Analysis of Parallel Scientific Applications

An integrated framework to profile, monitor, model and analyze power dissipation in parallel MPI and multi-threaded scientific applications and a statistical software module that inspects the execution trace of the application to calculate the parameters of an accurate model for the global energy consumption.

A new energy aware performance metric

This work argues what one should aim for is the total reduction of the spent energy in conjunction with minimization of time to solution, and proposes to use f(timetosolution)⋅energy (FTTSE) as the performance metric.

Ecient Sparse Matrix-Vector Multiplication on CUDA

Data structures and algorithms for SpMV that are eciently implemented on the CUDA platform for the ne-grained parallel architecture of the GPU and develop methods to exploit several common forms of matrix structure while oering alternatives which accommodate greater irregularity are developed.

Mixed Precision Iterative Refinement Techniques for the Solution of Dense Linear Systems

By using a combination of 32-bit and 64-bit floating point arithmetic, the performance of many dense and sparse linear algebra algorithms can be significantly enhanced while maintaining the 64-bit

Toward a New Metric for Ranking High Performance Computing Systems

A new high performance conjugate gradient (HPCG) benchmark is described, composed of computations and data access patterns more commonly found in applications that strive for a better correlation to real scientific application performance.