arXiv subjects

Stefan Zellmann

Publications and source records attributed to Stefan Zellmann.

At least 19 recordsLinked to original sources

arXiv2026

Streami: An MPI Data-Parallel Library to Compute Field Lines on GPUs

We present Streami, an extensible GPU-accelerated library for the computation of field lines in fluid flows on high-performance computers. Streami acts as a thin layer used for both post-hoc or in-situ analysis and can interface with existing MPI applications. We discuss Streami's application programming interface, key design decisions that led to Streami's high performance and extensibility, as well as extensions to support different fluid flow field representations. We also present a sample application for rapid prototyping and interactive seed point placement. Streami is released under a permissive open-source software license.

cs.CE↗

arXiv2026

RAFI -- A Ray/Work Forwarding Infrastructure for Data Parallel Multi-Node/Multi-GPU Computing

We present RaFI, a CUDA and MPI based software framework that simplifies the task of building GPU-enabled data-parallel software where rays or similar work items need to migrate between different GPUs. RaFI provides a simple interface for CUDA kernels to forward such work items to other GPUs, while under the hood managing all the CUDA and MPI related work required to make this happen. We describe RaFI's motivation and implementation, and show its potential in several example applications.

cs.DC↗

arXiv2025

GPU Volume Rendering with Hierarchical Compression Using VDB

We propose a compression-based approach to GPU rendering of large volumetric data using OpenVDB and NanoVDB. We use OpenVDB to create a lossy, fixed-rate compressed representation of the volume on the host, and use NanoVDB to perform fast, low-overhead, and on-the-fly decompression during rendering. We show that this approach is fast, works well even in a (incoherent) Monte Carlo path tracing context, can significantly reduce the memory requirements of volume rendering, and can be used as an almost drop-in replacement into existing 3D texture-based renderers.

cs.GR↗

arXiv2025

Data Parallel Visualization and Rendering on the RAMSES Supercomputer with ANARI

3D visualization and rendering in HPC are very heterogenous applications, though fundamentally the tasks involved are well-defined and do not differ much from application to application. The Khronos Group's ANARI standard seeks to consolidate 3D rendering across sci-vis applications. This paper makes an effort to convey challenges of 3D rendering and visualization with ANARI in the context of HPC, where the data does not fit within a single node or GPU but must be distributed. It also provides a gentle introduction to parallel rendering concepts and challenges to practitioners from the field of HPC in general. Finally, we present a case study showcasing data parallel rendering on the new supercomputer RAMSES at the University of Cologne.

cs.GR↗

arXiv2024

Standardized Data-Parallel Rendering Using ANARI

We propose and discuss a paradigm that allows for expressing \emph{data-parallel} rendering with the classically non-parallel ANARI API. We propose this as a new standard for data-parallel sci-vis rendering, describe two different implementations of this paradigm, and use multiple sample integrations into existing apps to show how easy it is to adopt this paradigm, and what can be gained from doing so.

cs.GR↗

arXiv2023

Beyond ExaBricks: GPU Volume Path Tracing of AMR Data

Adaptive Mesh Refinement (AMR) is becoming a prevalent data representation for scientific visualization. Resulting from large fluid mechanics simulations, the data is usually cell centric, imposing a number of challenges for high quality reconstruction at sample positions. While recent work has concentrated on real-time volume and isosurface rendering on GPUs, the rendering methods used still focus on simple lighting models without scattering events and global illumination. As in other areas of rendering, key to real-time performance are acceleration data structures; in this work we analyze the major bottlenecks of data structures that were originally optimized for camera/primary ray traversal when used with the incoherent ray tracing workload of a volumetric path tracer, and propose strategies to overcome the challenges coming with this.

cs.GR↗

arXiv2023

A Practical Guide to Implementing Off-Axis Stereo Projection Using Existing Ray Tracing Libraries

Virtual reality (VR) renderers driving CAVEs and similar immersive environments use the off-axis stereo camera model so that a tracked user can move freely in front of the projection plane. Geometrically, off-axis projection results in asymmetric viewing frusta and generalizes the ubiquitous perspective camera model to support positioning off the center of the projection plane. VR renderers often integrate with larger visualization systems that rely on libraries for position tracking and pose estimates, for ray tracing-based rendering, and for user interaction. We demonstrate different strategies to implement off-axis stereo projection within the constraints of given VR applications and ray tracing libraries. We aim for minimal to no adjustments required to the internal camera representation of such libraries. We include host and shader code with the article that can be directly integrated in custom applications.

cs.GR↗

arXiv2023

Immersive ExaBrick: Visualizing Large AMR Data in the CAVE

Rendering large adaptive mesh refinement (AMR) data in real-time in virtual reality (VR) environments is a complex challenge that demands sophisticated techniques and tools. The proposed solution harnesses the ExaBrick framework and integrates it as a plugin in COVISE, a robust visualization system equipped with the VR-centric OpenCOVER render module. This setup enables direct navigation and interaction within the rendered volume in a VR environment. The user interface incorporates rendering options and functions, ensuring a smooth and interactive experience. We show that high-quality volume rendering of AMR data in VR environments at interactive rates is possible using GPUs.

cs.GR↗

arXiv2023

Visual Analysis of Large Multi-Field AMR Data on GPUs Using Interactive Volume Lines

To visually compare ensembles of volumes, dynamic volume lines (DVLs) represent each ensemble member as a 1D polyline. To compute these, the volume cells are sorted on a space-filling curve and scaled by the ensemble's local variation. The resulting 1D plot can augment or serve as an alternative to a 3D volume visualization free of visual clutter and occlusion. Interactively computing DVLs is challenging when the data is large, and the volume grid is not structured/regular, as is often the case with computational fluid dynamics simulations. We extend DVLs to support large-scale, multi-field adaptive mesh refinement (AMR) data that can be explored interactively. Our GPU-based system updates the DVL representation whenever the data or the alpha transfer function changes. We demonstrate and evaluate our interactive prototype using large AMR volumes from astrophysics simulations.

cs.GR↗

arXiv2022

KiloNeuS: A Versatile Neural Implicit Surface Representation for Real-Time Rendering

NeRF-based techniques fit wide and deep multi-layer perceptrons (MLPs) to a continuous radiance field that can be rendered from any unseen viewpoint. However, the lack of surface and normals definition and high rendering times limit their usage in typical computer graphics applications. Such limitations have recently been overcome separately, but solving them together remains an open problem. We present KiloNeuS, a neural representation reconstructing an implicit surface represented as a signed distance function (SDF) from multi-view images and enabling real-time rendering by partitioning the space into thousands of tiny MLPs fast to inference. As we learn the implicit surface locally using independent models, resulting in a globally coherent geometry is non-trivial and needs to be addressed during training. We evaluate rendering performance on a GPU-accelerated ray-caster with in-shader neural network inference, resulting in an average of 46 FPS at high resolution, proving a satisfying tradeoff between storage costs and rendering quality. In fact, our evaluation for rendering quality and surface recovery shows that KiloNeuS outperforms its single-MLP counterpart. Finally, to exhibit the versatility of KiloNeuS, we integrate it into an interactive path-tracer taking full advantage of its surface normals. We consider our work a crucial first step toward real-time rendering of implicit neural representations under global illumination.

cs.CV↗

arXiv2022

GPU-based Data-parallel Rendering of Large, Unstructured, and Non-convexly Partitioned Data

Computational fluid dynamic simulations often produce large clusters of finite elements with non-trivial, non-convex boundaries and uneven distributions among compute nodes, posing challenges to compositing during interactive volume rendering. Correct, in-place visualization of such clusters becomes difficult because viewing rays straddle domain boundaries across multiple compute nodes. We propose a GPU-based, scalable, memory-efficient direct volume visualization framework suitable for in~situ and post~hoc usage. Our approach reduces memory usage of the unstructured volume elements by leveraging an exclusive or-based index reduction scheme and provides fast ray-marching-based traversal without requiring large external data structures built over the elements themselves. Moreover, we present a GPU-optimized deep compositing scheme that allows correct order compositing of intermediate color values accumulated across different ranks that works even for non-convex clusters. Our method scales well on large data-parallel systems and achieves interactive frame rates during visualization. We can interactively render both Fun3D Small Mars Lander (14 GB / 798.4 million finite elements) and Huge Mars Lander (111.57 GB / 6.4 billion finite elements) data sets at 14 and 10 frames per second using 72 and 80 GPUs, respectively, on TACC's Frontera supercomputer.

cs.GR↗

arXiv2022

Successful Common Envelope Ejection and Binary Neutron Star Formation in 3D Hydrodynamics

A binary neutron star merger has been observed in a multi-messenger detection of gravitational wave (GW) and electromagnetic (EM) radiation. Binary neutron stars that merge within a Hubble time, as well as many other compact binaries, are expected to form via common envelope evolution. Yet five decades of research on common envelope evolution have not yet resulted in a satisfactory understanding of the multi-spatial multi-timescale evolution for the systems that lead to compact binaries. In this paper, we report on the first successful simulations of common envelope ejection leading to binary neutron star formation in 3D hydrodynamics. We simulate the dynamical inspiral phase of the interaction between a 12$M_\odot$ red supergiant and a 1.4$M_\odot$ neutron star for different initial separations and initial conditions. For all of our simulations, we find complete envelope ejection and final orbital separations of $a_{\rm f} \approx 1.3$-$5.1 R_\odot$ depending on the simulation and criterion, leading to binary neutron stars that can merge within a Hubble time. We find $α_{\rm CE}$-equivalent efficiencies of $\approx 0.1$-$2.7$ depending on the simulation and criterion, but this may be specific for these extended progenitors. We fully resolve the core of the star to $\lesssim 0.005 R_\odot$ and our 3D hydrodynamics simulations are informed by an adjusted 1D analytic energy formalism and a 2D kinematics study in order to overcome the prohibitive computational cost of simulating these systems. The framework we develop in this paper can be used to simulate a wide variety of interactions between stars, from stellar mergers to common envelope episodes leading to GW sources.

astro-ph.HE↗

arXiv2022

Volkit: A Performance-Portable Computer Vision Library for 3D Volumetric Data

We present volkit, an open source library with high performance implementations of image manipulation and computer vision algorithms that focus on 3D volumetric representations. Volkit implements a cross-platform, performance-portable API targeting both CPUs and GPUs that defers data and resource movement and hides them from the application developer using a managed API. We use volkit to process medical and simulation data that is rendered in VR and consequently integrated the library into the C++ virtual reality software CalVR. The paper presents case studies and performance results and by that demonstrates the library's effectiveness and the efficiency of this approach.

cs.CV↗

arXiv2022

Point Containment Queries on Ray Tracing Cores for AMR Flow Visualization

Modern GPUs come with dedicated hardware to perform ray/triangle intersections and bounding volume hierarchy (BVH) traversal. While the primary use case for this hardware is photorealistic 3D computer graphics, with careful algorithm design scientists can also use this special-purpose hardware to accelerate general-purpose computations such as point containment queries. This article explains the principles behind these techniques and their application to vector field visualization of large simulation data using particle tracing.

cs.GR↗

arXiv2020

Ray Tracing Structured AMR Data Using ExaBricks

Structured Adaptive Mesh Refinement (Structured AMR) enables simulations to adapt the domain resolution to save computation and storage, and has become one of the dominant data representations used by scientific simulations; however, efficiently rendering such data remains a challenge. We present an efficient approach for volume- and iso-surface ray tracing of Structured AMR data on GPU-equipped workstations, using a combination of two different data structures. Together, these data structures allow a ray tracing based renderer to quickly determine which segments along the ray need to be integrated and at what frequency, while also providing quick access to all data values required for a smooth sample reconstruction kernel. Our method makes use of the RTX ray tracing hardware for surface rendering, ray marching, space skipping, and adaptive sampling; and allows for interactive changes to the transfer function and implicit iso-surfacing thresholds. We demonstrate that our method achieves high performance with little memory overhead, enabling interactive high quality rendering of complex AMR data sets on individual GPU workstations.

cs.GR↗

arXiv2020

Accelerating Force-Directed Graph Drawing with RT Cores

Graph drawing with spring embedders employs a V x V computation phase over the graph's vertex set to compute repulsive forces. Here, the efficacy of forces diminishes with distance: a vertex can effectively only influence other vertices in a certain radius around its position. Therefore, the algorithm lends itself to an implementation using search data structures to reduce the runtime complexity. NVIDIA RT cores implement hierarchical tree traversal in hardware. We show how to map the problem of finding graph layouts with force-directed methods to a ray tracing problem that can subsequently be implemented with dedicated ray tracing hardware. With that, we observe speedups of 4x to 13x over a CUDA software implementation.

cs.DS↗

arXiv2020

Augmenting Image Warping-Based Remote Volume Rendering with Ray Tracing

We propose an image warping-based remote rendering technique for volumes that decouples the rendering and display phases. Our work builds on prior work that samples the volume on the client using ray casting and reconstructs a z-value based on some heuristic. The color and depth buffer are then sent to the client that reuses this depth image as a stand-in for subsequent frames by warping it according to the current camera position until new data was received from the server. We augment that method by implementing the client renderer using ray tracing. By representing the pixel contributions as spheres, this allows us to effectively vary their footprint based on the distance to the viewer, which we find to give better results than point-based rasterization when applied to volumetric data sets.

cs.GR↗

arXiv2019

Adding Custom Intersectors to the C++ Ray Tracing Template Library Visionaray

Most ray tracing libraries allow the user to provide custom functionality that is executed when a potential ray surface interaction was encountered to determine if the interaction was valid or traversal should be continued. This is e.g. useful for alpha mask validation and allows the user to reuse existing ray object intersection routines rather than reimplementing them. Augmenting ray traversal with custom intersection logic requires some kind of callback mechanism that injects user code into existing library routines. With template libraries, this injection can happen statically since the user compiles the binary code herself. We present an implementation of this "custom intersector" approach and its integration into the C++ ray tracing template library Visionaray.

cs.GR↗