The authors present here the description of the main features of MOLFLOW+, a new software package that allows the calculation of several physical quantities of interest to vacuum engineers and scientists, such as pressure profiles, effective pumping speeds, adsorption distributions, angle of incidence or effusion profiles, and more. MOLFLOW+ is a follow up to the code MOLFLOW, which has been developed by one of us since 1991, and used for the analysis and design of many vacuum systems and components. MOLFLOW+ makes use of modern trends such as OpenGL graphic interface, multicore processors, graphical processing units (GPUs), and runs under different operating systems. In addition to implementing the test-particle Monte Carlo (TPMC) algorithm, as done previously in MOLFLOW, MOLFLOW+ also implements an alternate algorithm, the angular coefficient (AC) method. This article goes into some details of the TPMC and AC methods as implemented in MOLFLOW+, shows results of benchmark runs and comparisons with published numerical and analytical data, and discusses the advantages and disadvantages of the two methods. CUDA is the compute unified device architecture, a C++-like development environment for a class of GPUs. (C) 2009 American Vacuum Society. [DOI: 10.1116/1.3153280]