ID

E*

${\text{PERM}}_{{t}_{exp}}$

REMC_{
pm
}

REMC_{
m
}


L01

4

4 (< 1 sec)

4 (< 1 sec)

4 (< 1 sec)

L02

8

8 (< 1 sec)

8 (< 1 sec)

8 (< 1 sec)

L03

12

12 (< 1 sec)

12 (< 1 sec)

12 (< 1 sec)

L04

16

16 (32 sec)

16 (7 sec)

16 (5 sec)

L05

20

20 (3 hrs†)

20 (1.1 min)

20 (55 sec)

L06

24

23

24 (16 min)

24 (13 min)

L07

28

26 (33 sec)

28 (3.2 hrs)

28 (2.5 hrs)

L08

32

30 (3 min)

32 (50 hrs)

32 (16 hrs)

L09

36

34 (22 min)

35 (99 hrs)

35 (100 hrs)

L010

40

38 (40 min)

38 (9.6 hrs)

39 (100 hrs)

 The L0structures proposed in [52] are hard for both REMC and PERM to fold. After L05, PERM is unable to find the unique optimal conformation in any of the 100 independent runs conducted. REMC is unable to find the groundstate conformation for L09 and L010, however, REMC_{
m
}finds better suboptimal conformations than PERM in both instances. When only one folding direction of PERM finds the optimal conformation, we report the mean runtime of that direction, denoting this in the table with a †. When best energies found by ${\text{PERM}}_{{t}_{1}}$ and ${\text{PERM}}_{{t}_{2}}$ differed (and neither find the optimal solution), the best energy by either is reported and the runtime is omitted. For every instance, 100 independent runs were conducted of 1 CPU hour each. In cases where not every run reached the same energy value after 1 hour, the expected runtime to reach the energy value shown in the table was calculated using the equation detailed by Parkes and Walser [54].