Learn More
We present a methodology which allows to derive accurate and simple models which are able to describe the performance of parallel applications without looking at the source code. A trace is obtained and linear models are derived by fitting the outcome of a set of simulations varying the influential parameters, such as: processor speed, network latency or(More)
BACKGROUND This study examines how changes in cognition over time are related to participation in social activities and the extent of social networks. METHODS Data are drawn from a population-based, longitudinal study that began in 1989 among elderly Taiwanese. An over-dispersed Poisson model is used to regress the number of failed cognitive tasks (0-5)(More)
In the context of developing next-generation high-performance computing systems, there is often a need for an " end-to-end " simulation tool that can simulate the behaviour of a full application on a reasonably faithful model of the actual system. Considering the ever-increasing levels of paral-lelism, we take a communication-centric view of the system(More)
—Dragonfly networks have been recently proposed for the interconnection network of forthcoming exascale supercom-puters. Relying on large-radix routers, they build a topology with low diameter and high throughput, divided into multiple groups of routers. While minimal routing is appropriate for uniform traffic patterns, adversarial traffic patterns can(More)
We describe a methodology to derive a simple characterization of a parallel program and models of its performance on a target architecture. Our approach starts from an instrumented run of the program to obtain a trace. A simple linear model of the performance of the application as a function of architectural parameters is then derived by fitting the results(More)
The personalized all-to-all collective exchange is one of the most challenging communication patterns in HPC applications in terms of performance and scalability. In the context of the fat tree family of interconnection networks, widely used in current HPC systems and datacenters, we show that there is potential for optimizing this traffic pattern by(More)