Cody Rivera
University of Alabama
9 Papers
33 Citations
Cody Rivera is an academic researcher from University of Alabama. The author has contributed to research in topics: CUDA & Computer science. The author has an hindex of 4, co-authored 8 publications.
Chat about Author
Papers
cuSZ: An Efficient GPU-Based Error-Bounded Lossy Compression Framework for Scientific Data
Jiannan Tian,Sheng Di,Kai Zhao,Cody Rivera,Megan Hickman Fulp,Robert Underwood,Sian Jin,Xin Liang,Jon Calhoun,Dingwen Tao,Franck Cappello +10 more
TL;DR: This paper presents an optimized GPU version of the SZ compressor, cuSZ, and proposes a dual-quantization scheme to entirely remove the data dependency in the prediction step of SZ such that this step can be performed very efficiently on GPUs.
TSM2X: High-performance tall-and-skinny matrix–matrix multiplication on GPUs
TL;DR: This paper proposes two efficient algorithms---TSM2R and TSM2L---for two classes of tall-and-skinny matrix-matrix multiplications on GPUs that focus on optimizing linear algebra operation with at least one of the input matrices is tall- and- Skinny.
30
cuSZ: An Efficient GPU-Based Error-Bounded Lossy Compression Framework for Scientific Data
Jiannan Tian,Sheng Di,Kai Zhao,Cody Rivera,Megan Hickman Fulp,Robert Underwood,Sian Jin,Xin Liang,Jon Calhoun,Dingwen Tao,Franck Cappello +10 more
- 30 Sep 2020
TL;DR: In this article, a dual-quantization scheme is proposed to eliminate the data dependency in the prediction step of SZ, which can be performed very efficiently on GPUs and improves the compression ratio by up to 3.48x on the tested data.
Revisiting Huffman Coding: Toward Extreme Performance on Modern GPU Architectures
Jiannan Tian,Cody Rivera,Sheng Di,Jieyang Chen,Xin Liang,Dingwen Tao,Franck Cappello +6 more
- 17 May 2021
TL;DR: Zhang et al. as mentioned in this paper proposed and implemented an efficient Huffman encoding approach based on modern GPU architectures, which addresses two key challenges: (1) how to parallelize the entire Huffman decoding algorithm, including codebook construction, and (2) how fully utilize the high memory-bandwidth feature of modern GPU architecture.
•Posted Content
Revisiting Huffman Coding: Toward Extreme Performance on Modern GPU Architectures
TL;DR: This paper proposes and implements an efficient Huffman encoding approach based on modern GPU architectures, which addresses two key challenges: how to parallelize the entire Huffman encode algorithm, including codebook construction, and how to fully utilize the high memory-bandwidth feature of modern GPU architecture.