Improved performance of QCD code onALiCE

Improved performance of QCD code on ALiCE

In a typical lattice QCD project the total runtime of code on a supercomputing platform is often measured in months or even years. This means that even a modest improvement in the performance of the code can yield very tangible benefits. There are two aspects to the optimisation of code for parallel machines: single-node optimisation and the minimisation of the overhead incurred by inter-node c...

A Code Reordering Transformation for Improved Cache Performance

We prove theorems that show that if we can reorder a program's memory reference stream such that the reordered memory reference stream satis es a disjointness property, then the transformed program corresponding to the reordered stream is guaranteed to have fewer misses for any cache with arbitrary size or organization so long as the cache uses the LRU replacement policy. We can apply these res...

Platform independent profiling of a QCD code

The supercomputing platforms available for high performance computing based research evolve at a great rate. However, this rapid development of novel technologies requires constant adaptations and optimizations of the existing codes for each new machine architecture. In such context, minimizing time of efficiently porting the code on a new platform is of crucial importance. A possible solution ...

QCD Thermodynamics with Improved Actions

O(a) Improved Lattice Qcd *

We review the O(a) improvement of lattice QCD with special emphasis on the motivation for performing the improvement programme non-perturbatively and the general concepts of on-shell improvement. The present status of the calculations of various improvement coefficients (perturbative and non-perturbative) is reviewed, as well as the computation of the isospin current normalization constants ZA ...