NVIDIA GTC is set to resume following the COVID-19 pandemic, marking the return of the in-person GTC experience. OpenZeka is thrilled to participate in the exhibition, presenting its products to customers and visitors alike. The event serves as an invaluable opportunity to witness the latest technology and NVIDIA developments, while also networking with potential business partners. GTC stands as the premier AI conference, boasting a vast array of workshops encompassing topics such as HPC & AI, Large Language Models, Computer Vision, text and data analytics, NLP, graphics, and Digital Twins. Attendees can also engage in live keynotes delivered by NVIDIA CEO and Founder, Jensen Huang, making it a not-to-be-missed event for anyone in the AI community.
The OpenZeka team is gearing up to showcase solutions at the upcoming exhibition, all powered by NVIDIA technologies. Among our offerings, we’ll be featuring Cordatus AI, Generative AI, Omniverse, and Robotics solutions. These solutions highlight the versatility and power of NVIDIA’s technology stack. We look forward to sharing these exciting developments with attendees and exploring potential collaborations in this dynamic space.
Cordatus
Cordatus AI is a dynamic platform geared towards AI-driven video analytics and device management, offering a tailored solution for constructing a custom-made, enterprise-level AI computing environment. This versatile platform supports deployment across edge, on-premises, and cloud infrastructures.
Cordatus AI is an Intelligent Video Analytics (IVA) program designed to analyze raw camera streams, providing real-time and actionable insights to optimize businesses across various sectors. Its core function involves establishing seamless connectivity between sensors, such as cameras, and designated computing resources. By efficiently routing data from sensors to computing nodes, Cordatus AI facilitates integration, ensuring smooth operations.
Additionally, the platform offers robust device monitoring and management and application orchestrating capabilities, empowering users to control the entire system and utilize resources optimally. Cordatus leverages NVIDIA’s hardware and software ecosystems, operating on NVIDIA Jetson Platforms and x86-based systems powered by NVIDIA dGPUs. By utilizing NVIDIA’s DeepStream SDK,Video Codec SDK, CUDA, cuDNN, Triton Inference Server, Tensor RT and TAO Toolkit, the system enhances its performance and efficiency, delivering advanced capabilities and operations.
Generative AI
Generative AI represents a significant advancement in artificial intelligence, enabling the creation of new and diverse content across various domains. Its ability to automate creative tasks and tackle complex challenges has revolutionized innovation in unprecedented ways. By enhancing computer vision technologies with diverse and high-quality training data, Generative AI greatly improves their ability to recognize and process visual information, thus driving innovation and expanding the capabilities of machine vision systems.
Understanding the pivotal role of Generative AI, the OpenZeka team has integrated two sample applications from NVIDIA Generative AI Playground into the Cordatus AI as inference engines. This integration enables users to implement both ready-to-run and custom Generative AI applications seamlessly within the Cordatus AI, fully utilizing its comprehensive features for enhanced outcomes. The integration of NanoOWL and Live LLaVA Generative AI applications into the Cordatus AI platform stands as a testament to this approach. Users can now effortlessly operate and benefit from the functionalities offered by both engines within the Cordatus AI orchestration environment. With the NanoOWL application, we aim to showcase the efficient execution of real-time object search based on a multimodal Generative AI system, with a special focus on identifying objects that are traditionally challenging to detect. Concurrently, the Live LLaVA engine will provide users the ability to ask questions based on the live scene within a stream and receive real-time answers related to the video content they are examining. This entire setup is designed to operate seamlessly on the NVIDIA Jetson AGX ORIN platform.
Omniverse and Robotics
OpenZeka is highly interested in the possibilities presented by NVIDIA Omniverse and robotics. The emergence of 3D worlds and the concept of digital twins compel us to reimagine traditional problem-solving approaches. Leveraging NVIDIA technologies, OpenZeka is dedicated to developing innovative solutions through the integration of digital twins and robotics.
One of our ongoing initiatives involves the development of autonomous robots for disaster relief operations. In this project, we utilize an array of NVIDIA stack solutions including Jetson Orin,NVIDIA Omniverse, Isaac Sim, and Omniverse Replicator.
This synergistic approach enables us to explore new frontiers in simulation, collaboration, and automation, ultimately driving forward progress and innovation in various industries.