SK hynix Invests in Semidynamics to Develop Memory-Centric AI Chips

Barcelona-based Semidynamics announced a strategic investment from SK hynix to jointly develop next-generation AI infrastructure optimized for memory-intensive inference workloads, announced in a press release.

The collaboration aims to align processor and memory architectures to improve efficiency in large-scale AI systems. Semidynamics designs its processors around the open RISC-V architecture, focusing on overcoming the memory wall through its proprietary Gazzillion memory subsystem, which reduces data movement bottlenecks in AI inference.

The company recently completed a 3nm silicon tape-out with TSMC, one of the first by a European semiconductor firm at that process node. The new funding will support additional tape-outs and system-level development, including rack platform buildout for data center-scale AI inference.

Semidynamics has previously secured €45 million in non-dilutive funding from European and Spanish innovation programs to advance its AI silicon and infrastructure technologies.