This study is motivated by the need to devise means to enhance heat transfer in configurations, like the back step, that appear in certain types of MEMS that involve fluid flow and that are not very efficient from the thermal transfer point of view. In particular, the work described in this paper studies the effect that a prescribed flow pulsation (defined by two control parameters: velocity pulsation frequency and pressure gradient amplitude at the inlet section) has on the heat transfer rate behind a backward facing step in the unsteady laminar 2-D regime. The working fluid that we have considered is water with temperature dependent viscosity and thermal conductivity. We have found that, for inlet pressure gradients that avoid flow reversal at both the upstream and downstream boundary conditions, the timeaveraged Nusselt number behind the step depends on the two above mentioned control parameters and is always larger than in the steady-state case. At Reynolds 100 and pulsating at the resonance frequency, the maximum time-averaged Nusselt number in the horizontal wall region located behind the step whose length is four times the step height is 55% larger than in the steady-case. Away from the resonant pulsation frequency, the time-averaged Nusselt number smoothly decreases and approaches its steady-state value.
In this paper, we present NanoCheckpoints which is a lightweight software-based checkpoint/restart scheme for taskparallel HPC applications. We leverage OmpSs, a task-based OpenMP derivative programming model (PM) and its Nanos asynchronous dataflow runtime. NanoCheckpoints achieves minimal overheads by checkpointing only tasks' inputs which are available for free in the OmpSs PM. We evaluate NanoCheckpoints by both pure task-parallel shared memory benchmarks (up to 16 cores) and hybrid OmpSs+MPI applications (up to 1024 cores). The results indicate that NanoCheckpoints has on average overhead 3% for shared memory benchmarks. The dataflow semantics of Nanos, where both checkpointing and error recovery are asynchronous, allows NanoCheckpoints to scale at large core counts even when high error rates are present. For hybrid OmpSs+MPI benchmarks, NanoCheckpoints has very low overhead, on average 2%, and high scalability.
ARTICLE INFO ABSTRACT
Keywords:Micro-heat sink Tip clearance Pressure dropThis article presents an experimental study on the optimisation of micro-heat sink configurations when both thermal effects and pressure drop are accounted fon The interest of the latter is that the practical engineering viability of some of these systems also depends on the required pumping power. The working ñuid was water and, according to typical power dissipation and system size requirements, the considered ñuid regime was either laminar or transitional, and not fully developed from the hydrodynamics point of view. Five configurations were considered: a reference geometry (selected for comparison purposes) made up of square section micro-channels, and four alternative configurations that involved the presence of a variable tip clearance in the design. The performance of the different configurations was compared with regard to both cooling efficiency and pressure drop. Finally, we also provide some practical guidelines for the engineering design of these types of systems.
In this work we propose partial task replication and check-pointing for task-parallel HPC applications to mitigate silent data corruption (SDC) errors. As the complete replication of all application tasks can be prohibitive due to resource costs, we introduce programmer-directed selective replication mechanism to provide fault-tolerance while decreasing costs. Results show that our scheme detects and corrects around 65% of SDC errors with only 4% overhead on average.Peer ReviewedPostprint (published version
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.