Sample Applications
A classic field for high quality 3D visualization is product design, especially in the automotive business. Before clay models are made or even prototypes are build, photo-realistic renderings of 3D CAD designs give designers a feedback about how the car will look like. Typically, such renderings take a significant amount of time and even when using large render farms for the task, it can take several minutes to create one still image.
The state of the art XtreemView ray-tracing kernels go beyond merely bringing this time-to-image latency down. With XtreemView, the very same photo-realistic images can be generated at interactive frame rates, giving product designers fundamentally new ways of experiencing their designs on the screen.
With the recent addition of high-quality path-tracing to the XtreemView engine, even more realistic pictures are possible now. Especially on surfaces, where ray-tracing typically falls short, such as dull, ambient materials, path-tracing adds beautiful depths of light distribution and illumination.
When it comes to visualizing large-scale 3D CAD models that consist of hundreds of millions of triangles, materials, and light sources, most visualization tools and libraries have to sacrifice quality for the sake of creating an image at all. The typical approach, loading all the triangle lists into the internal GPU only works up to a certain limit and beyond it becomes highly difficult.
XtreemView comes with a powerful polygon engine. Due to its parallel design, it is easy to add more compute power and memory if needed to visualize even the largest scenes and models. Up to billions of triangles, tetrahedrons, or hexahedrons are supported at interactive framerates.
In physics, simulations are an important part of understanding what’s going on in the world and to solve the problems that are related to it. Particle simulations are especially a helpful tool to understand the interaction between solid objects and fluids. Here the difficulty is updating the scene graph fast enough.
XtreemView has patented algorithms for its internal data structure that enable extremely fast scene traversal and collision detection. Even more revolutionary is the speed to rebuild those internal structures, when objects in the scene change shape or location.
Another perfect use case for the XtreemView technology is animated scenes in Gaming and/or Film. In such scenes, correct occlusion mapping of moving objects, collision detection between individual objects, as well as updating the underlying scene structure can be challenging to render software.
XtreemView has patented algorithms for its internal data structure that enable extremely fast scene traversal and collision detection. Even more revolutionary is the speed to rebuild those internal structures, when objects in the scene change shape or location.
Geophysical applications, especially in the oil&gas industry, are facing the daring task of visualizing terabytes of volume data in order to analyse subsurface structures and identifying new hydrocarbon deposits. By large, this task is accomplished today by looking at individual slices of datasets, sometimes even just parts of a full slice. This is not just inefficient as it doesn’t provide a ‘big picture’ kind of view, it’s also easy to miss information, by skipping too far ahead.
XtremView’s volume render kernels are the fastest in class and can render datasets of pretty much any size, as long as it fits into memory. And because of the parallel architecture, additional nodes can be added easily to increase the available memory to the required size. So you’ll never have to adjust your problem size to your hardware ever again, but now you can adjust the hardware size to your problem. And if you don’t have enough hardware, simply use the cloud!
