English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Tools annual

Global: Zendy Free Plan

Global: Zendy Tools monthly

Global: Zendy Plus monthly

As HPC systems continue to grow to exascale, the amount of data that needs tobe saved or transmitted is exploding. To this end, many previous works havestudied using error-bounded lossy compressors to reduce the data size andimprove the I/O performance. However, little work has been done for effectivelyoffloading lossy compression onto FPGA-based SmartNICs to reduce thecompression overhead. In this paper, we propose a hardware-algorithm codesignfor an efficient and adaptive lossy compressor for scientific data on FPGAs(called CEAZ), which is the first lossy compressor that can achieve highcompression ratios and throughputs simultaneously. Specifically, we propose anefficient Huffman coding approach that can adaptively update Huffman codewordsonline based on codewords generated offline, from a variety of representativescientific datasets. Moreover, we derive a theoretical analysis to support aprecise control of compression ratio under an error-bounded compression mode,enabling accurate offline Huffman codewords generation. This also helps uscreate a fixed-ratio compression mode for consistent throughput. In addition,we develop an efficient compression pipeline by adopting cuSZ'sdual-quantization algorithm to our hardware use cases. Finally, we evaluateCEAZ on five real-world datasets with both a single FPGA board and 128 nodes(to accelerate parallel I/O). Experiments show that CEAZ outperforms thesecond-best FPGA-based lossy compressor by 2.3X of throughput and 3.0X ofratio. It also improves MPI_File_write and MPI_Gather throughputs by up to28.9X and 37.8X, respectively.

CEAZ: Accelerating Parallel I/O via Hardware-Algorithm Co-Designed Adaptive Lossy Compression