Why this company
EXDYNA is building physical data infrastructure for Physical AI, embodied intelligence, world models and AI for Science. Its website frames the core product around FARMS: fast, AI-native, reconfigurable, modular and sustainable scientific computing that can generate real-world simulation data for next-generation foundation models.
Technical edge
The company aims to move 100-million to billion-cell scientific computing tasks from large supercomputer clusters toward local computing environments, reducing infrastructure, energy and memory burdens. Its modular AI-native scientific computing agent stack is relevant to real-world simulation, high-throughput dataset generation, fluid-structure interaction, multi-robot scenarios, city-scale wind-field simulation and humanoid robot training.
Why now
AI is moving beyond language and images into engineering, robotics and scientific discovery. These domains need systems that reason about physics, test possibilities and shorten iteration cycles. High-fidelity, scalable and efficient physical data generation can become a key layer for both next-generation models and industrial R&D.
Why we invested
EXDYNA sits at the intersection of three Zelos themes: AI infrastructure, research commercialization and tools that make complex industrial systems more efficient. Its Singapore-based team brings research depth, A*STAR ecosystem connectivity and membership in the NVIDIA Inception program.
How Zelos helps
We support overseas market development, strategic positioning and business-model definition, while connecting the company with research, industrial and capital networks. The goal is to help the team move from breakthrough technology to a repeatable commercial platform.
Future vision
EXDYNA has the potential to become an enabling layer for scientists, engineers and robotics teams building the next generation of intelligent physical systems.