Designing high-performance mpi libraries with on-the-fly compression for modern gpu clusters
While the memory bandwidth of accelerators such as GPU has significantly improved over
the last decade, the commodity networks such as Ethernet and InfiniBand are lagging in …
the last decade, the commodity networks such as Ethernet and InfiniBand are lagging in …
Accelerating mpi all-to-all communication with online compression on modern gpu clusters
Abstract As more High-Performance Computing (HPC) and Deep Learning (DL) applications
are adapting to scale using GPUs, the communication of GPU-resident data is becoming …
are adapting to scale using GPUs, the communication of GPU-resident data is becoming …
Data compression for climate data
The different rates of increase for computational power and storage capabilities of
supercomputers turn data storage into a technical and economical problem. Because …
supercomputers turn data storage into a technical and economical problem. Because …
MPC: a massively parallel compression algorithm for scientific data
A Yang, H Mukka, F Hesaaraki… - 2015 IEEE International …, 2015 - ieeexplore.ieee.org
Due to their high peak performance and energy efficiency, massively parallel accelerators
such as GPUs are quickly spreading in high-performance computing, where large amounts …
such as GPUs are quickly spreading in high-performance computing, where large amounts …
BurstZ+: Eliminating the communication bottleneck of scientific computing accelerators via accelerated compression
We present BurstZ+, an accelerator platform that eliminates the communication bottleneck
between PCIe-attached scientific computing accelerators and their host servers, via …
between PCIe-attached scientific computing accelerators and their host servers, via …
Accelerating lossy and lossless compression on emerging bluefield dpu architectures
Data compression has become a crucial technique in addressing performance bottlenecks
caused by increasing data volumes in High-Performance Computing (HPC), Big Data, and …
caused by increasing data volumes in High-Performance Computing (HPC), Big Data, and …
BurstZ: a bandwidth-efficient scientific computing accelerator platform for large-scale data
We present BurstZ, a bandwidth-efficient accelerator platform for scientific computing. While
accelerators such as GPUs and FPGAs provide enormous computing capabilities, their …
accelerators such as GPUs and FPGAs provide enormous computing capabilities, their …
Real-time synthesis of compression algorithms for scientific data
M Burtscher, H Mukka, A Yang… - SC'16: Proceedings of …, 2016 - ieeexplore.ieee.org
Many scientific programs produce large amounts of floating-point data that are saved for
later use. To minimize the storage requirement, it is worthwhile to compress such data as …
later use. To minimize the storage requirement, it is worthwhile to compress such data as …
Adaptive-compi: Enhancing mpi-based applications' performance and scalability by using adaptive compression
This paper presents an optimization of MPI communication, called Adaptive-CoMPI, based
on runtime compression of MPI messages exchanged by applications. The technique …
on runtime compression of MPI messages exchanged by applications. The technique …
Dynamic-CoMPI: Dynamic optimization techniques for MPI parallel applications
This work presents an optimization of MPI communications, called Dynamic-CoMPI, which
uses two techniques in order to reduce the impact of communications and non-contiguous …
uses two techniques in order to reduce the impact of communications and non-contiguous …