Nvidia’s new tool uses generative artificial intelligence to create 3D designs

Nvidia has launched a new tool called FlexiCubes, which is actually a new way of using artificial intelligence to generate 3D shapes for a wide range of 3D applications.

According to Hoshio, using FlexiCubes with physical simulations can make it easier to create flexible objects such as pretzels or sweets in 3D models.

Nvidia said their new method for creating 3D shapes using artificial intelligence, called FlexiCubes, has the potential to improve the efficiency and quality of producing these 3D shapes in a wide range of applications.

From scene reconstruction to building generative AI, the next generation of AI models have shown remarkable success in producing realistic and accurate 3D models. These models are often created as standard triangular shapes, which offer many advantages including compatibility with existing software, advanced hardware acceleration, and support for physical simulations.

However, while there are benefits to using artificial intelligence to generate 3D shapes, the quality of these shapes plays an important role in realizing these benefits. Nvidia researchers have developed FlexiCubes in a way that significantly improves the quality of 3D shapes created in 3D modeling processes.

Nvidia’s FlexiCubes method can help the artificial intelligence production process to create high-quality 3D shapes that accurately depict fine details in complex shapes. This is especially useful for processes that use photogrammetry (creating 3D models from photos) and artificial intelligence (creating new content using artificial intelligence).

Improved AI with better 3D shapes

The introduction of FlexiCubes brings significant improvements in the production processes of 3D shapes. The higher quality shapes produced by FlexiCubes excel at displaying intricate details, increasing the realism and overall accuracy of AI-generated 3D models. These shapes are particularly suitable for physical simulation, where the quality of the 3D shapes plays an important role in achieving efficient and robust simulations.

FlexiCubes also has a feature that allows it to generate tetrahedral shapes that are useful for physical simulations. This feature increases the potential applications of FlexiCubes in various industries.

Nvidia presented its latest research on FlexiCubes and other developments at the Siggraph 2023 conference in Los Angeles.