看到这句话了吗？

所有跟贴·加跟贴·新语丝读书论坛

送交者: wasguru 于 2009-01-06, 21:47:31:

回答: See this article 由 james_hussein_bond 于 2009-01-06, 21:38:48:

“even a negligibly small deliberate noise can cause irreversibility, or the loss of reversibility.”

都故意引入noise了，那当然是不可逆的了。

所有跟贴:

Unfortunately, the page talking about bit-MD is not visible - wasguru (244 bytes) 2009-01-08, 12:02:54 (291714)
Here is a book on google that has preview pages - james_hussein_bond (582 bytes) 2009-01-08, 06:58:11 (291677)
I don't have access to full text (无内容) - wasguru (0 bytes) 2009-01-07, 21:34:36 (291592)
If you simply used lattice MD, then that's true. - james_hussein_bond (226 bytes) 2009-01-07, 21:05:13 (291574)
Confining particles on lattice grids simply means rounding coordinates off. - wasguru (86 bytes) 2009-01-07, 20:36:24 (291560)
In the old days there used to be a method called lattice MD - james_hussein_bond (132 bytes) 2009-01-07, 20:29:28 (291558)
Discretization of what? time, or coordinates? - wasguru (72 bytes) 2009-01-07, 16:25:33 (291500)
According to them - james_hussein_bond (254 bytes) 2009-01-07, 16:10:21 (291489)
All right. Yes, for LJ(6,12), you don't have to. - wasguru (741 bytes) 2009-01-07, 15:42:40 (291471)
You never have to calculate square root if you use Leonard-Jones potential (无内容) - james_hussein_bond (0 bytes) 2009-01-07, 15:24:20 (291451)
The problem is, bit-reversible does not mean the simulation is exact - wasguru (639 bytes) 2009-01-07, 14:50:17 (291417)
Maybe not for you but evidently a lot of people (wink 008) - james_hussein_bond (130 bytes) 2009-01-07, 14:37:02 (291401)
这样的point，还用的着他去费力气验证？ - wasguru (64 bytes) 2009-01-07, 12:54:46 (291308)
That's the point (无内容) - james_hussein_bond (0 bytes) 2009-01-07, 05:40:17 (291168)

加跟贴