HPC in the UAE: 2026 Status in AI and Advanced Computing

This country update on HPC and AI in the United Arab Emirates is delivered by Horst Simon in a session chaired by Carla Osthoff of Brazil's National Laboratory of Scientific Computing. Simon outlines the UAE's national AI strategy, now several years old, and notes the country's number eight position in the Stanford AI Index. He focuses on G42, the conglomerate created in 2018 that underpins the UAE's national data cloud and AI strategy, with combined compute of 0.39 exaflops and five systems on the TOP500 list, two of which are physically installed in the United States pending export resolution. G42's expansion includes a UAE-US AI campus, gigawatt-scale datacenter plans, and a partnership with Cerebras, which operates Condor Galaxy systems in G42-managed US datacenters and recently announced a joint datacenter venture in India. He also covers the Technology Innovation Institute, creator of the Falcon models and among the first Arabic large language models; a planned research computing upgrade at Khalifa University with roughly 200 CPU and 80 GPU nodes; and ADNOC's AI plans built largely on Microsoft Azure and G42. Simon then describes ADIA Lab, the research lab funded by the Abu Dhabi Investment Authority that he leads, with application thrusts in climate science, health science, digital economy, and materials science built on trustworthy AI and HPC research, including a planned climate copilot combining satellite, sensor, and model data for the MENA region, and a digital economy RFP that drew 130 international proposals.