Molecular dynamics (MD) simulations have transformed our understanding of the nanoscale, driving breakthroughs in materials science, computational chemistry, and several other fields, including biophysics and drug design. Even on exascale supercomputers, however, runtimes are excessive for systems and timescales of scientific interest. Here, we demonstrate strong scaling of MD simulations on the Cerebras Wafer-Scale Engine. By dedicating a processor core for each simulated atom, we demonstrate a 457-fold improvement in timesteps per second versus the Frontier GPU-based Exascale platform, along with a large improvement in timesteps per unit energy. Reducing every year of runtime to less than a day unlocks currently inaccessible timescales of slow microstructure transformation processes that are critical for understanding material behavior and function.Our dataflow algorithm runs Embedded Atom Method (EAM) simulations at rates over 699k timesteps per second for problems with up to 800k atoms. This demonstrated performance is unprecedented for general-purpose processing cores.

The advancement of machine learning techniques and the heterogeneous architectures of most current supercomputers are propelling the demand for large multiscale simulations that can automatically and autonomously couple diverse components and map them to relevant resources to solve complex problems at multiple scales. Nevertheless, despite the recent progress in workflow technologies, current capabilities are limited to coupling two scales. In the first-ever demonstration of using three scales of resolution, we present a scalable and generalizable framework that couples pairs of models using machine learning and in situ feedback. We expand upon the massively parallel Multiscale Machine-Learned Modeling Infrastructure (MuMMI), a recent, award-winning workflow, and generalize the framework beyond its original design. We discuss the challenges and learnings in executing a massive multiscale simulation campaign that utilized over 600,000 node hours on Summit and achieved more than 98% GPU occupancy for more than 83% of the time. We present innovations to enable several orders of magnitude scaling, including simultaneously coordinating 24,000 jobs, and managing several TBs of new data per day and over a billion files in total. Finally, we describe the generalizability of our framework and, with an upcoming open-source release, discuss how the presented framework may be used for new applications.

Computational models can define the functional dynamics of complex systems in exceptional detail. However, many modeling studies face seemingly incommensurate requirements: to gain meaningful insights into some phenomena requires models with high resolution (microscopic) detail that must nevertheless evolve over large (macroscopic) length- and time-scales. Multiscale modeling has become increasingly important to bridge this gap. Executing complex multiscale models on current petascale computers with high levels of parallelism and heterogeneous architectures is challenging. Many distinct types of resources need to be simultaneously managed, such as GPUs and CPUs, memory size and latencies, communication bottlenecks, and filesystem bandwidth. In addition, robustness to failure of compute nodes, network, and filesystems is critical.

We have created a scalable implementation of the kinetic Monte-Carlo method, SPOCK (Scalable Parallel Optimistic Crystal Kinetics). Unlike most reported parallel implementations relying on approximation to achieve parallelism, our parallelization is exact and accomplished using the Time Warp paradigm. We demonstrate that our implementation exhibits near perfect scaling for two different and important classes of systems. It runs efficiently on Vulcan, a 24 thousand node BlueGene/Q machine, using all ~400 thousand cores and ~1.6 million MPI tasks. Further, we have run production simulations using the full Vulcan machine and requiring nearly all available system memory. In this paper we demonstrate these results, and discuss some important implementation details. The kinetic Monte-Carlo method is ubiquitous within the natural sciences, and important classes of problems have so far been limited to sequential simulation. For many scientific simulations, an exact parallel implementation of the kinetic Monte-Carlo method has the potential of being game changing.

The fully automatic generation of code that establishes the reversibility of arbitrary C/C++ code has been a target of research and engineering for more than a decade as reverse computation has become a central notion in large scale parallel discrete event simulation (PDES). The simulation models that are implemented for PDES are of increasing complexity and size and require various language features to support abstraction, encapsulation, and composition when building a simulation model. In this paper we focus on parallel simulation models that are written in C++ and present an approach and an evaluation for a fully automatically generated reversible code for a kinetic Monte-Carlo application implemented in C++. Although a significant runtime overhead is introduced with our technique, the assurance that the reverse code is generated automatically and correctly, is an enormous win that allows simulation model developers to write forward event code using the entire C++ language, and have that code automatically transformed into reversible code to enable parallel execution with the Rensselaer's Optimistic Simulation System (ROSS).

The potential impact of blood flow simulations on the diagnosis and treatment of patients suffering from vascular disease is tremendous. Empowering models of the full arterial tree can provide insight into diseases such as arterial hypertension and enables the study of the influence of local factors on global hemodynamics. We present a new, highly scalable implementation of the lattice Boltzmann method which addresses key challenges such as multiscale coupling, limited memory capacity and bandwidth, and robust load balancing in complex geometries. We demonstrate the strong scaling of a three-dimensional, high-resolution simulation of hemodynamics in the systemic arterial tree on 1,572,864 cores of Blue Gene/Q. Faster calculation of flow in full arterial networks enables unprecedented risk stratification on a perpatient basis. In pursuit of this goal, we have introduced computational advances that significantly reduce time-to-solution for biofluidic simulations.

We have developed a highly efficient and scalable cardiac electrophysiology simulation capability that supports groundbreaking resolution and detail to elucidate the mechanisms of sudden cardiac death from arrhythmia. We can simulate thousands of heartbeats at a resolution of 0.1 mm, comparable to the size of cardiac cells, thereby enabling scientific inquiry not previously possible. Based on scaling results from the partially deployed Sequoia IBM Blue Gene/Q machine at Lawrence Livermore National Laboratory and planned optimizations, we estimate that by SC12 we will simulate 8 -- 10 heartbeats per minute -- a time-to-solution 400 -- 500 times faster than the state-of-the-art. Performance between 8 and 11 PFlop/s on the full 1,572,864 cores is anticipated, representing 40 -- 55 percent of peak. The power of the model is demonstrated by illuminating the subtle arrhythmogenic mechanisms of anti-arrhythmic drugs that paradoxically increase arrhythmias in some patient populations.