Early experiences scaling VMD molecular visualization and analysis jobs on blue waters

Stone, John E.; Isralewitz, Barry; Schulten, Klaus

doi:10.1109/xsw.2013.10

Cited by 21 publications

(50 citation statements)

References 18 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…GPU memory capacity is a critical issue when working with petascale molecular dynamics trajectories produced on leadership class systems, and it must be directly addressed in the design of GPU algorithms. The Blue Waters XK7 GPU system software also supports hardware-accelerated OpenGL rasterization, presently a unique feature among petascale computers [20].…”

Section: Cray Xe6/xk7 Overviewmentioning

confidence: 99%

“…The ability to perform large scale rendering jobs directly on petascale supercomputers allows users to avoid the transfer of large simulation trajectories to other sites for rendering. The high performance parallel movie rendering infrastructure in VMD allows rapid completion of photorealistic movie rendering that would take weeks or months (even with the acceleration provided by GPUs) on a desktop workstation [20].…”

Section: Gpu Ray Tracingmentioning

confidence: 99%

“…Tachyon provides a lightweight software API that is easy to integrate, and, as a result of a decade of ongoing co-development with VMD, supports many rendering primitives, data types, and triangle mesh formats required for efficient rendering of molecular scenes produced by VMD. We have previously reported early results running the CPU-based versions of Tachyon on the Blue Waters Cray XE6/XK7 petascale computer, outlining performance optimizations and work-scheduling improvements that benefit both intra-node multi-core rendering performance as applied to VMD movie rendering jobs on large node counts [20].…”

Section: Gpu Ray Tracingmentioning

confidence: 99%

“…One issue that arose in rewriting Tachyon for the GPU was the limitation that GPU kernels are allowed to run only for a brief period of time before a so-called "watchdog" timer in the host operating system terminates the running kernel. This behavior is required by most windowing systems and arises on Blue Waters as a result of its support for off-screen OpenGL Pbuffer rendering [20]. In practice, this prevents ray tracing kernels from being written traditionally with significant internal looping for shooting shadow test rays, ambient occlusion samples, transmission rays or reflection rays, because the runtime for a ray launch can exceed the timeout limit.…”

Section: Gpu Ray Tracingmentioning

confidence: 99%

“…The incorporation of GPUs into petascale computing systems has created a new opportunity to apply GPU acceleration to the creation of advanced visualizations and movie renderings. Over the past year VMD has been adapted for execution on petascale systems using a hybrid computational approach that combines distributed memory message passing with MPI, shared memory multithreading, and GPU acceleration [20]. We describe the ongoing adaptation of VMD to exploit the unique visualization capabilities of the Cray XK7 GPU compute nodes, including GPU-accelerated algorithms for molecular surface visualization adapted to handle very large biomolecular complexes, and a new GPU-accelerated ray tracing engine capable of reproducing complex shadowing and lighting effects not feasible in interactive visualization tools or rasterization APIs such as OpenGL.…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

GPU-accelerated molecular visualization on petascale supercomputing platforms

Stone

Vandivort

Schulten

2013

Proceedings of the 8th International Workshop on Ultrascale Visualization

Self Cite

View full text Add to dashboard Cite

Petascale supercomputers create new opportunities for the study of the structure and function of large biomolecular complexes such as viruses and photosynthetic organelles, permitting all-atom molecular dynamics simulations of tens to hundreds of millions of atoms. Together with simulation and analysis, visualization provides researchers with a powerful "computational microscope". Petascale molecular dynamics simulations produce tens to hundreds of terabytes of data that can be impractical to transfer to remote facilities, making it necessary to perform visualization and analysis tasks in-place on the supercomputer where the data are generated. We describe the adaptation of key visualization features of VMD, a widely used molecular visualization and analysis tool, for GPU-accelerated petascale computers. We discuss early experiences adapting ray tracing algorithms for GPUs, and compare rendering performance for recent petascale molecular simulation test cases on Cray XE6 (CPUonly) and XK7 (GPU-accelerated) compute nodes. Finally, we highlight opportunities for further algorithmic improvements and optimizations.

show abstract

Section: Cray Xe6/xk7 Overviewmentioning

confidence: 99%

Section: Gpu Ray Tracingmentioning

confidence: 99%

Section: Gpu Ray Tracingmentioning

confidence: 99%

Section: Gpu Ray Tracingmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

GPU-accelerated molecular visualization on petascale supercomputing platforms

Stone

Vandivort

Schulten

2013

Proceedings of the 8th International Workshop on Ultrascale Visualization

Self Cite

View full text Add to dashboard Cite

show abstract

Parallel performance of molecular dynamics trajectory analysis

Khoshlessan

Paraskevakos

Fox

et al. 2020

Concurrency and Computation

View full text Add to dashboard Cite

The performance of biomolecular molecular dynamics (MD) simulations has steadily increased on modern high performance computing (HPC) resources but acceleration of the analysis of the output trajectories has lagged behind so that analyzing simulations is becoming a bottleneck.To close this gap, we studied the performance of parallel trajectory analysis with MPI and the Python MDAnalysis library on three different XSEDE supercomputers where trajectories were read from a Lustre parallel file system. Strong scaling performance was impeded by stragglers, MPI processes that were slower than the typical process. Stragglers were less prevalent for compute-bound workloads, thus pointing to file reading as a crucial bottleneck for scaling. However, a more complicated picture emerged in which both the computation and the ingestion of data exhibited close to ideal strong scaling behavior whereas stragglers were primarily caused by either large MPI communication costs or long times to open the single shared trajectory file. We improved overall strong scaling performance by either subfiling (splitting the trajectory into separate files) or MPI-IO with Parallel HDF5 trajectory files. We obtained near ideal strong scaling on up to 384 cores (16 nodes), thus reducing trajectory analysis times by two orders of magnitude compared to the serial approach.

show abstract

Interactive Ray Tracing Techniques for High-Fidelity Scientific Visualization

Stone

2019

Ray Tracing Gems

Self Cite

View full text Add to dashboard Cite

This chapter describes rendering techniques and implementation considerations when using ray tracing for interactive scientific and technical visualization. Ray tracing offers a convenient framework for building high-fidelity rendering engines that can directly generate publication-quality images for scientific manuscripts while also providing high interactivity in a what-you-see-is-what-you-get rendering experience. The combination of interactivity with sophisticated rendering enables scientists who are typically not experts in computer graphics or rendering technologies to be able to immediately apply advanced rendering features in their daily work. This chapter summarizes techniques and practical approaches learned from applying ray tracing techniques to scientific visualization, and molecular visualization in particular. 27.1 INTRODUCTION Scientific and technical visualizations are used to illustrate complex data, concepts, and physical phenomena to aid in the development of hypotheses, discover design problems, facilitate collaboration, and inform decision making. The scenes that arise in such visualizations incorporate graphical representations of the details of key structures and mechanisms and their relationships, or the dynamics of complex processes under study. High-quality ray tracing techniques have been of great use in the creation of visualizations that elucidate complex scenes. Interactivity is a powerful aid to the effectiveness of scientific visualization because it allows the visualization user to rapidly explore and manipulate data, models, and graphical representations to obtain insights and to help confirm or deny hypotheses.

show abstract

Early experiences scaling VMD molecular visualization and analysis jobs on blue waters

Cited by 21 publications

References 18 publications

GPU-accelerated molecular visualization on petascale supercomputing platforms

GPU-accelerated molecular visualization on petascale supercomputing platforms

Parallel performance of molecular dynamics trajectory analysis

Interactive Ray Tracing Techniques for High-Fidelity Scientific Visualization

Contact Info

Product

Resources

About