Quantum Monte Carlo by message passing

Conference · Sat May 01 04:00:00 EDT 1993

OSTI ID:10155064

Bonca, J; Gubernatis, J E

We summarize results of quantum Monte Carlo simulations of the degenerate single-impurity Anderson model using the impurity algorithm of Hirsch and Fye. Using methods of Bayesian statistical inference, coupled with the principle of maximum entropy, we extracted the single-particle spectral density from the imaginary-time Green`s function. The variations of resulting spectral densities with model parameters agree qualitatively with the spectral densities predicted by NCA calculations. All the simulations were performed on a cluster of 16 IBM R6000/560 workstations under the control of the message-passing software PVM. We described the trivial parallelization of our quantum Monte Carlo code both for the cluster and the CM-5 computer. Other issues for effective parallelization of the impurity algorithm are also discussed.

• Research Organization: Los Alamos National Lab., NM (United States)
• Sponsoring Organization: USDOE, Washington, DC (United States)
• DOE Contract Number: W-7405-ENG-36
• OSTI ID: 10155064
• Report Number(s): LA-UR--93-1296; CONF-9302112--2; ON: DE93012645
• Country of Publication: United States
• Language: English

Related Subjects

665000
75 CONDENSED MATTER PHYSICS
SUPERCONDUCTIVITY AND SUPERFLUIDITY
99 GENERAL AND MISCELLANEOUS
990200
ALGORITHMS
COMPUTERIZED SIMULATION
DILUTE ALLOYS
GREEN FUNCTION
HAMILTONIANS
KONDO EFFECT
MAGNETIC MATERIALS
MAGNETIC MOMENTS
MAGNETIC SUSCEPTIBILITY
MATHEMATICS AND COMPUTERS
MONTE CARLO METHOD
PARALLEL PROCESSING
PHYSICS OF CONDENSED MATTER