@InProceedings{8902635,
  author = {F. Cabeleira and A. Mariano and G. Falcao},
  booktitle = {2019 27th European Signal Processing Conference (EUSIPCO)},
  title = {Memory-Optimized Voronoi Cell-based Parallel Kernels for the Shortest Vector Problem on Lattices},
  year = {2019},
  pages = {1-5},
  abstract = {In this paper we propose a parallel implementation of a Voronoi cell-based algorithm for the Shortest Vector Problem for both CPU and GPU architectures. Additionally, we present an algorithmic simplification with particular emphasis on significantly reducing the memory usage of the implementation. According to our tests, the parallel multi-core CPU implementation scales linearly with the number of cores used, and also benefits from simultaneous multi-threading, achieving a maximum speedup of 5.56× for 8 threads. The parallel GPU implementation obtains speedups of 13.08×, compared with the sequential CPU implementation. The acceleration of this class of signal processing algorithms is a fundamental step in the evolution of post-quantum cryptanalysis. Currently, the best algorithms can take months to process for moderately low dimensions.},
  keywords = {computational geometry;graphics processing units;multiprocessing systems;multi-threading;vectors;Voronoi cell-based algorithm;shortest vector problem;simultaneous multithreading;parallel GPU implementation;sequential CPU implementation;signal processing algorithms;memory-optimized Voronoi cell-based parallel kernels;multicore CPU implementation;post-quantum cryptanalysis;Signal processing algorithms;Lattices;Graphics processing units;Instruction sets;Cryptography;Computer architecture;Kernel;Cryptography;Voronoi;Accelerators},
  doi = {10.23919/EUSIPCO.2019.8902635},
  issn = {2076-1465},
  month = {Sep.},
  url = {https://www.eurasip.org/proceedings/eusipco/eusipco2019/proceedings/papers/1570529174.pdf},
}

{"_id":"mTTs9eENFtcLgnA8T","bibbaseid":"cabeleira-mariano-falcao-memoryoptimizedvoronoicellbasedparallelkernelsfortheshortestvectorproblemonlattices-2019","authorIDs":[],"author_short":["Cabeleira, F.","Mariano, A.","Falcao, G."],"bibdata":{"bibtype":"inproceedings","type":"inproceedings","author":[{"firstnames":["F."],"propositions":[],"lastnames":["Cabeleira"],"suffixes":[]},{"firstnames":["A."],"propositions":[],"lastnames":["Mariano"],"suffixes":[]},{"firstnames":["G."],"propositions":[],"lastnames":["Falcao"],"suffixes":[]}],"booktitle":"2019 27th European Signal Processing Conference (EUSIPCO)","title":"Memory-Optimized Voronoi Cell-based Parallel Kernels for the Shortest Vector Problem on Lattices","year":"2019","pages":"1-5","abstract":"In this paper we propose a parallel implementation of a Voronoi cell-based algorithm for the Shortest Vector Problem for both CPU and GPU architectures. Additionally, we present an algorithmic simplification with particular emphasis on significantly reducing the memory usage of the implementation. According to our tests, the parallel multi-core CPU implementation scales linearly with the number of cores used, and also benefits from simultaneous multi-threading, achieving a maximum speedup of 5.56× for 8 threads. The parallel GPU implementation obtains speedups of 13.08×, compared with the sequential CPU implementation. The acceleration of this class of signal processing algorithms is a fundamental step in the evolution of post-quantum cryptanalysis. Currently, the best algorithms can take months to process for moderately low dimensions.","keywords":"computational geometry;graphics processing units;multiprocessing systems;multi-threading;vectors;Voronoi cell-based algorithm;shortest vector problem;simultaneous multithreading;parallel GPU implementation;sequential CPU implementation;signal processing algorithms;memory-optimized Voronoi cell-based parallel kernels;multicore CPU implementation;post-quantum cryptanalysis;Signal processing algorithms;Lattices;Graphics processing units;Instruction sets;Cryptography;Computer architecture;Kernel;Cryptography;Voronoi;Accelerators","doi":"10.23919/EUSIPCO.2019.8902635","issn":"2076-1465","month":"Sep.","url":"https://www.eurasip.org/proceedings/eusipco/eusipco2019/proceedings/papers/1570529174.pdf","bibtex":"@InProceedings{8902635,\n author = {F. Cabeleira and A. Mariano and G. Falcao},\n booktitle = {2019 27th European Signal Processing Conference (EUSIPCO)},\n title = {Memory-Optimized Voronoi Cell-based Parallel Kernels for the Shortest Vector Problem on Lattices},\n year = {2019},\n pages = {1-5},\n abstract = {In this paper we propose a parallel implementation of a Voronoi cell-based algorithm for the Shortest Vector Problem for both CPU and GPU architectures. Additionally, we present an algorithmic simplification with particular emphasis on significantly reducing the memory usage of the implementation. According to our tests, the parallel multi-core CPU implementation scales linearly with the number of cores used, and also benefits from simultaneous multi-threading, achieving a maximum speedup of 5.56× for 8 threads. The parallel GPU implementation obtains speedups of 13.08×, compared with the sequential CPU implementation. The acceleration of this class of signal processing algorithms is a fundamental step in the evolution of post-quantum cryptanalysis. Currently, the best algorithms can take months to process for moderately low dimensions.},\n keywords = {computational geometry;graphics processing units;multiprocessing systems;multi-threading;vectors;Voronoi cell-based algorithm;shortest vector problem;simultaneous multithreading;parallel GPU implementation;sequential CPU implementation;signal processing algorithms;memory-optimized Voronoi cell-based parallel kernels;multicore CPU implementation;post-quantum cryptanalysis;Signal processing algorithms;Lattices;Graphics processing units;Instruction sets;Cryptography;Computer architecture;Kernel;Cryptography;Voronoi;Accelerators},\n doi = {10.23919/EUSIPCO.2019.8902635},\n issn = {2076-1465},\n month = {Sep.},\n url = {https://www.eurasip.org/proceedings/eusipco/eusipco2019/proceedings/papers/1570529174.pdf},\n}\n\n","author_short":["Cabeleira, F.","Mariano, A.","Falcao, G."],"key":"8902635","id":"8902635","bibbaseid":"cabeleira-mariano-falcao-memoryoptimizedvoronoicellbasedparallelkernelsfortheshortestvectorproblemonlattices-2019","role":"author","urls":{"Paper":"https://www.eurasip.org/proceedings/eusipco/eusipco2019/proceedings/papers/1570529174.pdf"},"keyword":["computational geometry;graphics processing units;multiprocessing systems;multi-threading;vectors;Voronoi cell-based algorithm;shortest vector problem;simultaneous multithreading;parallel GPU implementation;sequential CPU implementation;signal processing algorithms;memory-optimized Voronoi cell-based parallel kernels;multicore CPU implementation;post-quantum cryptanalysis;Signal processing algorithms;Lattices;Graphics processing units;Instruction sets;Cryptography;Computer architecture;Kernel;Cryptography;Voronoi;Accelerators"],"metadata":{"authorlinks":{}},"downloads":0},"bibtype":"inproceedings","biburl":"https://raw.githubusercontent.com/Roznn/EUSIPCO/main/eusipco2019url.bib","creationDate":"2021-02-11T19:15:21.942Z","downloads":0,"keywords":["computational geometry;graphics processing units;multiprocessing systems;multi-threading;vectors;voronoi cell-based algorithm;shortest vector problem;simultaneous multithreading;parallel gpu implementation;sequential cpu implementation;signal processing algorithms;memory-optimized voronoi cell-based parallel kernels;multicore cpu implementation;post-quantum cryptanalysis;signal processing algorithms;lattices;graphics processing units;instruction sets;cryptography;computer architecture;kernel;cryptography;voronoi;accelerators"],"search_terms":["memory","optimized","voronoi","cell","based","parallel","kernels","shortest","vector","problem","lattices","cabeleira","mariano","falcao"],"title":"Memory-Optimized Voronoi Cell-based Parallel Kernels for the Shortest Vector Problem on Lattices","year":2019,"dataSources":["NqWTiMfRR56v86wRs","r6oz3cMyC99QfiuHW"]}