As the scale of high-performance computing (HPC) continues to grow, failure resilience of parallel applications becomes crucial. In this paper, we present FT-Pro, an adaptive fault...
We consider the impact of different communication architectures on the performability (performance + availability) of cluster-based servers. In particular, we use a combination of ...
The programming and retasking of sensor nodes could benefit greatly from the use of a virtual machine (VM) since byte code is compact, can be loaded on demand, and interpreted on...
Abstract. Many evolutionary algorithm applications involve either fitness functions with high time complexity or large dimensionality (hence very many fitness evaluations will typi...
— We present a tool, PerfCenter, that takes as input the deployment, configuration, message flow and workload details of the hardware and software servers in an application hos...