Eliyan, a Santa Clara-based developer of high-performance connectivity technology for AI and high-performance computing systems, announced it has secured $50 million in strategic investments from a group that includes AMD, Arm, Coherent, and Meta, alongside returning backers Samsung Catalyst Fund and Intel Capital.
The company said the financing reflects growing demand for power-efficient, high-bandwidth connectivity as AI architectures move beyond single packages and monolithic modules toward more disaggregated designs. As systems scale, Eliyan positioned its roadmap as spanning both on-package die-to-die links and off-package connectivity that extends across boards, systems, and racks.
Eliyan said the funding will support the manufacturing and qualification of its next-generation interconnect IP and chiplet products, expand ecosystem partnerships, and enable broader deployment across AI infrastructure, high-performance computing, and edge applications.
The company is advancing commercialization of its NuLink PHY and NuGear chiplet families, which it says are designed to address memory and I/O bottlenecks that can limit scalability, power efficiency, and cost in next-generation AI systems. Eliyan highlighted silicon-proven NuLink die-to-die technology at 64G, as well as next-generation chip-to-chip offerings that include single-ended 32G to 64Gbps and differential 224Gbps SerDes, with emerging 448G capabilities aimed at linking multiple packages and modules. It also pointed to NuGear chiplets focused on scale-up network connectivity, targeting link bandwidths from 1.6T to 12.8Tbps for demanding accelerator and memory expansion architectures, including environments that pair electrical links with optical engines.
The company also referenced its work aligned with emerging memory interconnect standards, including SPHBM4e, and described early silicon proof points that it says extend toward future SHHBM5.
KEY QUOTES
“This support from industry leaders who deploy AI infrastructure at massive scale underscores the importance of the innovation Eliyan is driving. Our silicon-proven 64G D2D leadership, which enables efficient on-package chiplet integration combined with our roadmap for 224G and 448G chip-to-chip interconnect, differentiated scale-up connectivity at 1.6T to 12.8T link bandwidths, and early SPHBM4e silicon proof points extending to future SHHBM5, positions Eliyan to enable the next generation of AI system architectures.”
Ramin Farjadrad, CEO and Co-Founder, Eliyan
“The evolution of AI systems increasingly depends on scalable, energy-efficient architectures and strong ecosystem collaboration. Arm is focused on enabling the technologies and partnerships required to support next-generation AI infrastructure at scale.”
Mohamed Awad, Executive Vice President, Cloud AI Business Unit, Arm
“Eliyan is addressing one of the most critical challenges in scaling AI and advanced computing: delivering efficient, high-bandwidth connectivity for chiplet-based architectures. This capability is foundational for the next generation of systems across data center, cloud, and edge markets, and we continue to recognize the value of Eliyan’s approach as they execute on this vision.”
Srini Ananth, Managing Director, Intel Capital
“AI infrastructure is growing at a dramatic pace, accelerating the need for high-performance connectivity technology to scale up compute and memory. We are very excited by Eliyan’s momentum since we led their prior round in 2024. The strong investors represented in this new financing round are a testament to the team’s strong execution across the PHY and chiplets businesses.”
Dede Goldschmidt, Senior Vice President, Managing Director and Head of Samsung Catalyst Fund
