Submission Details

Date 2022-04-06

Dimension 40

Relative error size α 0.035

Name(s) of Participant(s) Wang Leizhang
Xia Wenwen
Wang Geng
Gu Dawu
Wang Baocang

E-Mail Address wanggxx at sjtu.edu.cn

Algorithm improved progressive pnj-bkz(based on G6K)

Hardware CPU：AMD EPYC™ 7002 Series 128@2.6GHz RAM：512GB GPU：8 * NVIDIA®GeForce®RTX 3090 VRAM：8 * 24GB（936.2 GB/s） system：CentOS7.9

Runtime around 233.14 hours

Notes Thanks for the help of developer of G6K-GPU-Tensor for debugging the code!!\\The RAM cost is 184.4221 GB

Solution [1168 983 1142 1321 922 277 1432 1067 559 180 1427 209 706 697 313 980 83 1357 1130 1514 118 630 941 761 615 894 1554 1002 1407 660 263 523 553 1144 125 961 1427 1499 509 6]

Error norm 2324.6999806427