Aces4: A Platform for Computational Chemistry Calculations with Extremely Large Block-Sparse Arrays

Leveraging the Super Instruction Architecture to Develop Massively Parallel Computational and Environmental Chemistry Applications

“…Aces4, an open source redesign of the long standing community code ACESIII, is reaching the point of being ready for operational use, specially when there is a need for computing open-shell coupled cluster energies. The performance of Aces4 has surpassed its predecessor ACESIII, and the more flexible implementation of the SIA has already enabled new physics to be implemented [Sanders et al(2017)Sanders, Byrd, Jindal, Lotrich, Lyakh, Perera, and Bartlett].…”

“…In this section, we give a very brief overview of the SIA. It has been described in more detail, along with the quantum chemistry application Aces4 in [Sanders et al(2017) Sanders, Byrd, Jindal, Lotrich, Lyakh, Perera, and Bartlett], [Byrd et al(2018)Byrd, Bartlett, and Sanders].…”

“…Its original instantiation was ACE-SIII (Advanced Concepts in Electronic Structure) Lotrich, Flocke, Ponton, Yau, Perera, Deumens, and Bartlett], a software package for highly accurate ab initio electronic structure methods. Its successor, Aces4 [Sanders et al(2019) Sanders, Byrd, Jindal, Lotrich, Lyakh, Perera, and Bartlett], [Sanders et al(2017) Sanders, Byrd, Jindal, Lotrich, Lyakh, Perera, and Bartlett], [Byrd et al(2018)Byrd, Bartlett, and Sanders], was a complete from-the-ground-up reimplementation including extensions to the SIAL language, important new capabilities to the underling architecture such as support for block-sparse arrays, an improved software architecture, and implementations of new methodological and implementation developments enabled by the enhancements.…”

Domain‐specific virtual processors as a portable programming and execution model for parallel computational workloads on modern heterogeneous high‐performance computing architectures

“…In particular, the quantum many‐body formalism used in quantum chemistry heavily relies on large‐scale numerical linear and multi‐linear (tensor) algebra, thus making quantum‐chemistry applications highly attractive for massively parallel high‐performance computing (HPC) platforms. Numerous parallel quantum‐chemistry codes have been developed and widely adopted by computational chemistry community over the time . In particular, many of these codes were specifically designed to run on the distributed HPC systems based on multicore processors, following the stable hardware trend of 2000s.…”

“…In particular, we describe a concrete DSVP implementation, namely the tensor algebra virtual processor (TAVP) designed and implemented by the author in the form of the ExaTENSOR library . Although the TAVP microarchitecture has its own unique design introducing a number of novel elements such as the fully hierarchical hardware encapsulation, it can also be viewed as a generalization and evolution of earlier efforts, specifically the so‐called Super Instruction Architecture framework used in the ACES‐III and ACES‐IV software suites for expressing and executing quantum many‐body algorithms operating on large dense arrays of numbers. In this retrospective, DSVP is a variant of a Super Instruction Processor (SIP) on an abstract architectural level, but it differs from the previous concrete SIP implementations at the microarchitectural level, that is, its exposed implementation design is different.…”