posted Apr 20, 2011, 2:31 PM by Jeongnim Kim
posted Dec 1, 2010, 7:43 AM by Jeongnim Kim
"Fully accelerating quantum Monte Carlo simulations of real materials on GPU clusters," was accepted as a peer-review article in Computing in Science and Engineering. The preprint is available at CISE site. |
posted Nov 2, 2010, 8:11 AM by Jeongnim Kim
I tried to give PGI compilers another chance but they really make it hard.
What C++ compiler does not have std::abs(double)? Apparently, PGI doesn't.
more than one instance of overloaded function "std::abs" matches the argument list: function "abs(int)" function "std::abs(long)" function "std::abs(long double)" argument types are: (double)
|
posted Oct 3, 2010, 2:15 PM by Jeongnim Kim
[
updated Oct 3, 2010, 2:17 PM
]
posted Sep 25, 2010, 7:43 AM by Jeongnim Kim
[
updated Sep 25, 2010, 7:45 AM
]
posted Apr 3, 2010, 8:52 AM by Jeongnim Kim
posted Mar 2, 2010, 9:06 AM by Jeongnim Kim
[
updated Mar 2, 2010, 9:31 AM
]
posted Feb 23, 2010, 7:32 AM by Jeongnim Kim
A binary using r4648 is available on the project spaces for production runs.
On Jaguar /ccs/proj/mat001/xt5/qmcapp_xt5_r4648_20100223
On Kraken /nics/a/proj/qmc/bin/qmcapp_xt5_r4648_20100223
|
posted Feb 21, 2010, 8:12 AM by Jeongnim Kim
When no estimator is defined in a qmc section, LocalEnergyEstimator using ascii data format is used.
|
posted Feb 20, 2010, 5:28 AM by Jeongnim Kim
Ken's summary of the discussion on Feb/19.
1) Add a "wizard" utility to wfconv that will generate XML input with all possible parameters with reasonable defaults. 2) Fix as much as possible within QMCPACK itself, including aborting on inconsistent data, parameters outside the allowed range, etc. 3) Tightening up the parsing of attributes and parameters to balk at unrecognized parameters. 4) Introducing a "help" utility which gives possible parameters and attributes for elements. 5) Removing case sensitivity or agree on case conventions.
|
|