PA1:Fiduccia–Mattheyses algorithm (Partition)
Goal: Minimum two groups of cut-size
Benchmark
Pin num
Initial Cut_size
Final Cut_size
RunTime(s)
adapted1
210904
7963
7960
4.02
adapted2
254457
10068
10067
5.1
adapted3
450927
25809
25803
9.61
superbule1
765102
52750
52749
28.62
superbule1
921273
60973
60893
23.37
PA2:Simulated Annealing Algorithm (Placement)
Goal: Perturbation Randomly to place objects.
ami33_1_10 : H/W ratio 1 | white-space 10%
ami33_2_10 : H/W ratio 2 | white-space 10%
ami33_3_10 : H/W ratio 3 | white-space 10%
PA3:Improving Wirelength based on Abacus (Detail-Placement)
Goal: Improving Wirelength in 5 minute
============ SCL ==========
NumRows : 1829
Numrow_array : 1829
----------
============ NODE ==========
NumNodes : 847441
NumTerminals : 82339
Numnode_array : 847441
NumMovableNodes : 765102
NumTerminal : 52627
NumTerminal_NI : 29712
----------
============ LEGAL ==========
Numinitial_movable_node : 765102
NumInRow : 817729 (817729)
----------
============ SHAPE ==========
NumNonRectangularNodes : 37
NumNonRect : 37
----------
========== NET ==========
NumNets : 822744
NumPins : 2861188
Numnet_array : 822744
----------
Run time : 4.23 s
=============== LG ===============
============ ROW_LIST INITIAL ==========
row_counter : 817729
----------
============ TOTAL HPWL_CAL ==========
initial_HPWL : 339097654
total_HPWL : 339097654 (100 %)
----------
=============== ITER 0 ===============
Iter time : 60.31 s
=============== ITER 1 ===============
Iter time : 62.66 s
=============== ITER 2 ===============
Iter time : 63.58 s
=============== ITER 3 ===============
Iter time : 64.1 s
=============== ITER 4 ===============
Iter time : 39.66 s
============ TOTAL HPWL_CAL ==========
total_HPWL : 316232559 (93.2571 %)
----------
===============================
Run time : 296.1 s
PA4:Routing congestion estimation based on Flute (Routing)
Goal: Using Flute to transfer nets to rectilinear Steiner minimal trees (RSMT) and estimate routing congestion with them.
Ex: Benchmark adaptec1 (Routing blockage | Horizontal Routing Congestion | Vertical Routing Congestion)