AI Accelerators - Products Under Development

AI servers equipped with FPGA-based accelerators are designed for early deployment, validation, and workload evaluation, ensuring optimal performance alongside advanced GPU architecture.

AI servers that integrate FPGA-based accelerators and leverage GPU architecture are designed for handling data centre-scale AI workloads.

Rack-scale AI infrastructure featuring AI servers is designed for scalable deployment in enterprise and sovereign data centers, utilizing FPGA-based accelerators and advanced GPU architecture.

A next-generation GPU architecture is currently under development to support large-scale AI servers for training and inference, leveraging FPGA-based accelerators for enhanced performance.

Chip-based accelerator systems, including FPGA-based accelerators, are also being designed in parallel, with the 1st test chip, Dirac-1, currently under development for AI servers and advancements in GPU architecture.