Understanding GPU-Based Lossy Compression for Extreme-Scale Cosmological Simulations

Jin, Sian; Grosset, Pascal; Biwer, C.; Pulido, Jesus; Tian, Jiannan; Tao, Dingwen; Ahrens, James

doi:10.1109/ipdps47924.2020.00021

Cited by 42 publications

(35 citation statements)

References 26 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We compare cuSZ with cuZFP in terms of the kernel performance and the overall performance that includes the GPU-to-CPU communication cost. Note that the performance of cuZFP is highly related to its userset fixed bitrate according to the previous study [39] Throughput ( / ) performance of cuSZ is hardly affected by the user-set error bound. Therefore, we choose the acceptable fixed bitrate for cuZFP, which generates data distortion (i.e., PSNR of about 85 dB) similar to that of cuSZ, as shown in Table 5.…”

Section: Compression Performancementioning

confidence: 61%

“…can be applied to decompression, as mentioned in §3.3. Here we argue that the compression throughput is more important than the decompression throughput, because users use the CPU-SZ mainly to decompress the data for postanalysis and visualization instead of the GPU after the compressed data is transferred and stored to parallel file systems [11,39]. We note that cuSZ on the CESM-ATM dataset exhibits much lower performance than on other datasets.…”

Section: Compression Performancementioning

confidence: 90%

“…Recently, the ZFP team released their CUDA version, called cuZFP [15]. cuZFP provides much higher throughputs for compression and decompression compared with the CPU version [39]. However, the current cuZFP only supports fixed-rate mode, which significantly limit its adoption in practice.…”

Section: Related Work 51 Gpu-accelerated Scientific Compressionmentioning

confidence: 99%

See 2 more Smart Citations

cuSZ

Tian

Zhao

et al. 2020

Proceedings of the ACM International Conference on Parallel Architectures and Compilation Techniques

Self Cite

View full text Add to dashboard Cite

Error-bounded lossy compression is a state-of-the-art data reduction technique for HPC applications because it not only significantly reduces storage overhead but also can retain high fidelity for postanalysis. Because supercomputers and HPC applications are becoming heterogeneous using accelerator-based architectures, in particular GPUs, several development teams have recently released GPU versions of their lossy compressors. However, existing state-of-the-art GPU-based lossy compressors suffer from either low compression and decompression throughput or low compression quality. In this paper, we present an optimized GPU version, cuSZ, for one of the best error-bounded lossy compressors-SZ. To the best of our knowledge, cuSZ is the first error-bounded lossy compressor on GPUs for scientific data. Our contributions are fourfold. (1) We propose a dual-qantization scheme to entirely remove the data dependency in the prediction step of SZ such that this step can be performed very efficiently on GPUs. (2) We develop an efficient customized Huffman coding for the SZ compressor on GPUs. (3) We implement cuSZ using CUDA and optimize its performance by improving the utilization of GPU memory bandwidth. (4) We evaluate our cuSZ on five real-world HPC application datasets from the Scientific Data Reduction Benchmarks and compare it with other state-of-the-art methods on both CPUs and GPUs. Experiments show that our cuSZ improves SZ's compression throughput by up to 370.1× and 13.1×, respectively, over the production version running on single and multiple CPU cores, respectively, while getting the same quality of reconstructed data. It also improves the compression ratio by up to 3.48× on the tested data compared with another state-of-the-art GPU supported lossy compressor.

show abstract

Section: Compression Performancementioning

confidence: 61%

Section: Compression Performancementioning

confidence: 90%

See 1 more Smart Citation

cuSZ

Tian

Zhao

et al. 2020

Proceedings of the ACM International Conference on Parallel Architectures and Compilation Techniques

Self Cite

View full text Add to dashboard Cite

show abstract

“…On the other hand, using the new generation of error-bounded lossy compression techniques, such as SZ [9,22,34] and ZFP [23], we can achieve much higher compression ratios with minimal distortion of the data as demonstrated in many prior studies [4, 9, 16, 21-23, 26, 28, 34, 35]. However, previous approaches of utilizing lossy compression for scientific datasets have always applied the same compression configuration to the entire dataset [21,33]. Yet, if we look at a visualization of baryon density in a Nyx simulation, shown in Figure 1, we can see that not all partitions (regions) have the same amount of information.…”

Section: Introductionmentioning

confidence: 98%

“…Apart from fine-grained adaptive compression, we must also be able to precisely control the compression error for domain-specific post-hoc analysis. Research has shown that general-purpose data distortion metrics, such as peak signal-to-noise ratio (PSNR), normalized root-mean-square error, mean relative error (MRE), and mean square error (MSE), on their own cannot satisfy the demand of quality for cosmological simulation post-hoc analysis [16,21]. For example, PSNR does not tell us how the mass of a halo would be impacted after compression.…”

Section: Introductionmentioning

confidence: 99%

Adaptive Configuration of In Situ Lossy Compression for Cosmology Simulations via Fine-Grained Rate-Quality Modeling

Jin

Pulido

Grosset

et al. 2021

Proceedings of the 30th International Symposium on High-Performance Parallel and Distributed Computing

Self Cite

View full text Add to dashboard Cite

Extreme-scale cosmological simulations have been widely used by today's researchers and scientists on leadership supercomputers. A new generation of error-bounded lossy compressors has been used in workflows to reduce storage requirements and minimize the impact of throughput limitations while saving large snapshots of high-fidelity data for post-hoc analysis. In this paper, we propose to adaptively provide compression configurations to compute partitions of cosmological simulations with newly designed postanalysis aware rate-quality modeling. The contribution is fourfold:(1) We propose a novel adaptive approach to select feasible error bounds for different partitions, showing the possibility and efficiency of adaptively configuring lossy compression for each partition individually. (2) We build models to estimate the overall loss of post-analysis result due to lossy compression and to estimate compression ratio, based on the property of each partition.(3) We develop an efficient optimization guideline to determine the best-fit configuration of error bounds combination in order to maximize the compression ratio under acceptable post-analysis quality loss. (4) Our approach introduces negligible overheads for feature extraction and error-bound optimization for each partition, enabling post-analysis-aware in situ lossy compression for cosmological simulations. Experiments show that our proposed models are highly accurate and reliable. Our fine-grained adaptive configuration approach improves the compression ratio of up to 73% on the tested datasets with the same post-analysis distortion with only 1% performance overhead.

show abstract

Fulfilling the Promises of Lossy Compression for Scientific Applications

Cappello

Gok

2020

Communications in Computer and Information Science

View full text Add to dashboard Cite

Understanding GPU-Based Lossy Compression for Extreme-Scale Cosmological Simulations

Cited by 42 publications

References 26 publications

cuSZ

cuSZ

Adaptive Configuration of In Situ Lossy Compression for Cosmology Simulations via Fine-Grained Rate-Quality Modeling

Fulfilling the Promises of Lossy Compression for Scientific Applications

Contact Info

Product

Resources

About