2019R2-Structures.pdf

节选段落一：
• Improved scaling for SMART Fracture simulations (TurboBlade model)
0
2
4
6
8
10
4 8 16 32 64 128 256
Sp
ee
d
u
p
Number of Cores
DMP Scaling Performance
2019 R1
2019 R2
• 2.7 million DOF; sparse solver
• Nonlinear static analysis involving
fracture with 1 crack (roughly 100
crack tip nodes)
• Linux


节选段落二：
cluster; each compute node
contains 2 Intel Xeon Gold 6142
processors (32 cores), 384GB RAM,
SSD, CentOS 7.4
Crack
Distributed ANSYS Enhancements
• Improved scaling for Normal Lagrange contact (Model 1)
0
2
4
6
8
10
12
2 4 8 16 32 64
Sp
ee
d
u
p
Number of Cores
DMP Scaling Performance
R19.1
R19.2
2019


节选段落三：
Lagrange contact (Model 2)
0
2
4
6
8
10
12
14
16
18
2 4 8 16 32 64 128 256
Sp
ee
d
u
p
Number of Cores
DMP Scaling Performance
R19.1
R19.2
2019 R1
2019 R2
• 5.9 MDOF; sparse solver
• Nonlinear static analysis involving
Normal Lagrange contact and CE/CP
• Linux cluster; each compute node
contains 2 Intel