The proceedings contain 96 papers. The topics discussed include: taking a quantum leap in time to solution for simulations of high-Tc superconductors;20 petaflops simulation of protein suspensions in crowding conditio...
ISBN:
(纸本)9781450323789
The proceedings contain 96 papers. The topics discussed include: taking a quantum leap in time to solution for simulations of high-Tc superconductors;20 petaflops simulation of protein suspensions in crowding conditions;radiative signatures of the relativistic Kelvin-Helmholtz instability;HACC: extreme scaling and performance across diverse architectures;ACR: automatic checkpoint/restart for soft and hard error protection;SPBC: leveraging the characteristics of MPI HPC applications for scalable checkpointing;using simulation to explore distributed key-value stores for extreme-scale system services;a large-scale cross-architecture evaluation of thread-coarsening;a framework for load balancing of tensor contraction expressions via dynamic task partitioning;distributed wait state tracking for runtime MPI deadlock detection;globalizing selectively: shared-memory efficiency with address-space separation;and performance evaluation of Intel transactional synchronization extensions for highperformancecomputing.
The proceedings contain 6 papers. The topics discussed include: an approximate thermal-aware Q-routing for optical NoCs;on the feasibility of hybrid electrical/optical switch architecture for large-scale training of d...
ISBN:
(纸本)9781728159812
The proceedings contain 6 papers. The topics discussed include: an approximate thermal-aware Q-routing for optical NoCs;on the feasibility of hybrid electrical/optical switch architecture for large-scale training of distributed deep learning;past, current and future technologies for optical submarine cables;simulations of photonic quantum networks for performanceanalysis and experiment design;an optical neural network architecture based on highly parallelized WDM-multiplier-accumulator;and scalable low-power high-performance rack-scale optical network.
The proceedings contain 9 papers. The topics discussed include: how the flang frontend works;an LLVM instrumentation plug-in for score-P;LLVM compiler implementation for explicit parallelization and SIMD vectorization...
ISBN:
(纸本)9781450355650
The proceedings contain 9 papers. The topics discussed include: how the flang frontend works;an LLVM instrumentation plug-in for score-P;LLVM compiler implementation for explicit parallelization and SIMD vectorization;implementing implicit OpenMP data sharing on GPUs;benchmarking and evaluating unified memory for OpenMP GPU offloading;PACXXv2 + RV an LLVM-based portable high-performance programming model;applying temporal blocking with a directive-based approach;and QUARC: an optimized DSL framework using LLVM.
The proceedings contain 8 papers. The topics discussed include: BatchFS: scaling the file system control plane with client-funded metadata servers;using property graphs for rich metadata management in hpc systems;eval...
ISBN:
(纸本)9781479970254
The proceedings contain 8 papers. The topics discussed include: BatchFS: scaling the file system control plane with client-funded metadata servers;using property graphs for rich metadata management in hpc systems;evaluating lustre's metadata server on a multi-socket platform;alleviating I/O interference via caching and rate-controlled prefetching without degrading migration performance;VSFS: a searchable distributed file system;automatic generation of I/O kernels for HPC applications;HPIS3: towards a high-performance simulator for hybrid parallel I/O and storage systems;and Feign: in-silico laboratory for researching I/O strategies.
The proceedings contain 64 papers. The topics discussed include: Egeria: a framework for automatic synthesis of HPC advising tools through multi-layered natural language processing;optimizing the query performance of ...
ISBN:
(纸本)9781450351140
The proceedings contain 64 papers. The topics discussed include: Egeria: a framework for automatic synthesis of HPC advising tools through multi-layered natural language processing;optimizing the query performance of block index through data analysis and I/O modeling;Sympiler: transforming sparse matrix codes by decoupling symbolic analysis;control replication: compiling implicit parallelism to efficient SPMD with logical regions;efficient process mapping in geo-distributed cloud data centers;topology-aware GPU scheduling for learning workloads in cloud environments;probabilistic guarantees of execution duration for amazon spot instances;a framework for scalable biophysics-based image analysis;and exploring and analyzing the real impact of modern on-package memory on HPC scientific kernels.
The proceedings contain 7 papers. The topics discussed include: Initial investigation of a scheme to use instantaneous CPU power consumption for energy savings;toward application-specific memory reconfiguration for en...
ISBN:
(纸本)9781450325042
The proceedings contain 7 papers. The topics discussed include: Initial investigation of a scheme to use instantaneous CPU power consumption for energy savings;toward application-specific memory reconfiguration for energy efficiency;modeling the effects of DFS on power consumption in hybrid chip multiprocessors;unified performance and power modeling of scientific workloads;leakage energy estimates for HPC applications;evaluating energy savings for checkpoint/restart;and OpenMP and MPI application energy measurement variation.
The proceedings contain 4 papers. The topics discussed include: abstract interpretation: testing at scale without testing at scale;climate models: challenges for Fortran development tools;experience with using a perfo...
ISBN:
(纸本)9781479970353
The proceedings contain 4 papers. The topics discussed include: abstract interpretation: testing at scale without testing at scale;climate models: challenges for Fortran development tools;experience with using a performance predictor during development: a distributed storage system tale;and towards test driven development for computational science with pFUnit.
The proceedings contain 10 papers. The topics discussed include: design and analysis of a 32-bit embedded high-performance cluster optimized for energy and performance;toward efficient programmer-managed two-level mem...
ISBN:
(纸本)9781479975648
The proceedings contain 10 papers. The topics discussed include: design and analysis of a 32-bit embedded high-performance cluster optimized for energy and performance;toward efficient programmer-managed two-level memory hierarchies in exascale computers;power profiling of a reduced data movement algorithm for neutron cross section data in Monte Carlo simulations;abstract machine models and proxy architectures for exascale computing;mPPM, viewed as a co-design effort;an evaluation of threaded models for a classical MD proxy application;performance and energy evaluation of CoMD on Intel Xeon Phi co-processors;application characterization using oxbow toolkit and PADS infrastructure;using a complementary emulation-simulation co-design approach to assess application readiness for processing-in-memory systems;and an implementation of block conjugate gradient algorithm on CPU-GPU processors.
The proceedings contain 10 papers. The topics discussed include: teaching parallel and distributed computing concepts in simulation with WRENCH;assessing the integration of parallel and distributed computing in early ...
ISBN:
(纸本)9781728159751
The proceedings contain 10 papers. The topics discussed include: teaching parallel and distributed computing concepts in simulation with WRENCH;assessing the integration of parallel and distributed computing in early undergraduate computer science curriculum using unplugged activities;successful systems in production graduate teaching;teaching concurrent and distributed programming with concepts over mathematical proofs;toward improving collaborative behaviour during competitive programming assignments;and teaching on demand: an HPC experience.
The proceedings contain 5 papers. The topics discussed include: towards scalable distributed graph database engine for hybrid clouds;to overlap or not to overlap: optimizing incremental MapReduce computations for on-d...
ISBN:
(纸本)9781479970346
The proceedings contain 5 papers. The topics discussed include: towards scalable distributed graph database engine for hybrid clouds;to overlap or not to overlap: optimizing incremental MapReduce computations for on-demand data upload;locality and network-aware reduce task scheduling for data-intensive applications;skyport - container-based execution environment management for multi-cloud scientific workflows;and integrating pig with harp to support iterative applications with fast cache and customized communication.
暂无评论