AI Played a Starring Role at SIGGRAPH 2024
By Jeff Berman, Editor
Advancements in artificial intelligence (AI) – especially generative AI – represented the key news announcements and technology demonstrations at @SIGGRAPH2024 at the Colorado Convention Center in Denver, Colorado, July 28-Aug. 1.
This was the first time the annual computer graphics conference was held in Colorado since its founding at The University of Colorado in Boulder 51 years ago.
Almost 9,000 industry professionals attended this time, according to the Association for Computing Machinery (ACM), which hosts the event. Attendees came from 76 countries, including the U.S., Canada, Japan, South Korea, France, the United Kingdom, Australia, Brazil and Nigeria.
When it came to news, most of the key takeaways came from the two keynote sessions featuring NVIDIA DIA CEO Jensen Huang, the second of which included him chatting with Meta CEO Mark Zuckerberg
In addition to providing fodder for many social media posts after the two executives swapped their jackets on stage, they discussed the transformative potential of open source AI and AI assistants, and Zuckerberg also announced the launch of AI Studio. The latter is a new Meta platform allowing users to create, share and discover AI characters, making AI more accessible to millions of creators and small businesses.
Other news announced at SIGGRAPH included: Autodesk unveiling the latest advances in generative AI and cloud-enabled workflows that it said help artists focus on creativity and empower teams to be more efficient.
Smart device brand OPPO highlighted AI research papers and introduced its latest AI phones, the Find X7 Ultra and Reno12 Pro, equipped with powerful GenAI capabilities.
IT solutions provider Supermicro announced a new addition to its SuperCluster portfolio of plug-and-play AI infrastructure solutions for the NVIDIA Omniverse platform to deliver high-performance, GenAI-enhanced 3D workflows at enterprise scale.
And Beamr Imaging and Oracle Cloud Infrastructure (OCI) demonstrated the optimization of high-quality, high-resolution video production rendered from 3D models. The Beamr Cloud video optimization service supported by OCI could offer a fast, easy solution to demanding tasks, Beamr said.
A demonstration that Beamr provided at SIGGRAPH showcased the optimization of a video from a 3D design, reducing a very large video file to one-fourth of its original size, the firm said. The streamlined video was from an OpenUSD-based scene developed on the NVIDIA Omniverse platform for developing and deploying 3D workflows, applications and services based on OpenUSD.
Conference presentations at SIGGRAPH also focused on the full scope of the impact of computer graphics and interactive techniques.
Other SIGGRAPH highlights included the presentation of more than 250 research papers; production sessions focusing on film, games and TV, including a look at the visual effects of the film Dune: Part Two and the animation in Kung Fu Panda 4; and a session with Google sharing its framework for its Project Starline technology project.
In Other News from SIGGRAPH
Autodesk returned to SIGGRAPH, unveiling the latest advances in generative AI and cloud-enabled workflows that help artists focus on creativity and empower teams to be more efficient.
In a demo at SIGGRAPH 2024, NVIDIA showcased the assistive 3D world-building capabilities of its Edify-powered models and the NVIDIA Omniverse platform.
IT solutions provider Supermicro announced a new addition to its SuperCluster portfolio of plug-and-play AI infrastructure solutions for the NVIDIA Omniverse platform to deliver high-performance, generative AI-enhanced 3D workflows at enterprise scale. The new SuperCluster features the latest Supermicro NVIDIA OVX systems and allows enterprises to easily scale as workloads increase, Supermicro said.
Updates that were announced at SIGGRAPH included a ChatRTX refresh and Llama 3 support to enhance AI content creation, productivity and development.
Introducing SAM 2: The next generation of Meta Segment Anything Model for videos and images (Meta Blog)
Following its success with the Meta Segment Anything Model (SAM) for images, the company announced during SIGGRAPH that it is releasing SAM 2, a unified model for real-time promptable object segmentation in images and videos that the company said “achieves state-of-the-art performance.”
Technicolor at SIGGRAPH 2024: Celebrating Innovation and Creativity
Technicolor’s presence at SIGGRAPH underscored its commitment to advancing the art and science of storytelling, leveraging cutting-edge technologies including AI to drive the future of entertainment, it said.
At SIGGRAPH, Beamr Imaging and Oracle Cloud Infrastructure (OCI) demonstrated the optimization of high-quality, high-resolution video production rendered from 3D models. The Beamr Cloud video optimization service supported by OCI may offer a fast, easy solution to demanding tasks, Beamr said. The demonstration at SIGGRAPH showcased the optimization of a video from a 3D design, reducing an extremely large video file to one-fourth of its original size. The streamlined video was from an OpenUSD-based scene developed on NVIDIA Omniverse platform for developing and deploying 3D workflows, applications and services based on OpenUSD.