G42 picks Qualcomm to boost AI inference performance

By Shilpa Annie Joseph, Official Reporter
  • Follow author on
G42 selects Qualcomm
Rep. Image

G42, the leading UAE-based technology holding group, has entered into a technology collaboration with Qualcomm Technologies to bring the high-performance, low-power Qualcomm Cloud AI 100 solutions to market through Core42’s Condor AI platform.

The Qualcomm Cloud AI 100 inference solutions are designed to provide industry-leading energy efficiency, portability, flexibility, and price performance for customers running computationally intense AI workloads.

Core42’s Condor AI is a cutting-edge cloud computing platform delivering an optimized infrastructure layer with solutions from market-leading AI providers. The Qualcomm Cloud AI 100 Ultra AI accelerators are engineered to deliver unparalleled inference performance and cost efficiency.

Kiril Evtimov, Group CTO of G42 and CEO of Core42, said that, “At G42, we collaborate with global technology leaders to explore new frontiers and redefine the potential of technology to create transformative opportunities for our partners, our customers, and society at large. With Qualcomm Cloud AI 100 Ultra, we will deliver unique access to unparalleled AI performance at a fraction of the current cost.”

“This solution enables global access to a platform tuned for efficient AI inference at cloud scale. This alliance delivers high-performance AI, easily deployed, at a cost that allows our customers to benefit from state-of-the-art model innovation,” stated Nakul Duggal, Group General Manager, Automotive, Industrial and Cloud, Qualcomm Technologies

By incorporating advanced data formats and novel AI techniques, such as unstructured sparsity, speculative sampling, and use-case specific NAS using the combination of Cerebras’ CS-3 and AI 100 Ultra, the Condor AI platform is designed to deliver up to 10x increase in tokens per dollar.

According to the statement, Core42’s Condor AI is delivering optimized production-grade solutions to both private and public cloud, allowing for the ability to ingest any trained model and deploy it for production.

Most Reads | Binance receives VASP license from Dubai’s VARA