Topic: Data Compression

Highlights include ML techniques for computed tomography, a scalable Gaussian process framework, safe and trustworthy AI, and autonomous multiscale simulations.

In two papers from the 2024 International Conference on Machine Learning, Livermore researchers investigate how LLMs perform under measurable scrutiny.

This issue highlights some of CASC’s contributions to the DOE's Exascale Computing Project.

zfp is an open-source C/C++ library for compressed floating-point and integer arrays that support high throughput read and write random access.

LLNL’s HPC capabilities play a significant role in international science research and innovation, and Lab researchers have won 10 R&D 100 Awards in the Software–Services category in the past decade.

Two LLNL teams have come up with ingenious solutions to a few of the more vexing difficulties. For their efforts, they’ve won awards coveted by scientists in the technology fields.

New research reveals subtleties in the performance of neural image compression methods, offering insights toward improving these models for real-world applications.

The Data and Visualization efforts in the DOE’s Exascale Computing Project provide an ecosystem of capabilities for data management, analysis, lossy compression, and visualization.

With this year’s results, the Lab has now collected a total of 179 R&D 100 awards since 1978. The awards will be showcased at the 61st R&D 100 black-tie awards gala on Nov. 16 in San Diego.

LLNL's zfp and Variorum software projects are winners. LLNL is a co-developing organization on the winning CANDLE project.

Unique among data compressors, zfp is designed to be a compact number format for storing data arrays in-memory in compressed form while still supporting high-speed random access.

High-precision numerical data from computer simulations, observations, and experiments is often represented in floating point and can easily reach terabytes to petabytes of storage.

LLNL is participating in the 34th annual Supercomputing Conference (SC22), which will be held both virtually and in Dallas on November 13–18, 2022.

Winning the best paper award at PacificVis 2022, a research team has developed a resolution-precision-adaptive representation technique that reduces mesh sizes, thereby reducing the memory and storage footprints of large scientific datasets.

fpzip is a library for lossless or lossy compression of multidimensional floating-point arrays. It was primarily designed for lossless compression.

Highlights include the directorate's annual external review, machine learning for ALE simulations, CFD modeling for low-carbon solutions, seismic modeling, and an in-line floating point compression tool.

hzip 1.0.1 is a C++ library for lossless compression of structured and unstructured meshes composed of cells with hypercube topology.