Cornelis, a technology firm focused on high-performance networking for artificial intelligence, announced on Monday that it has successfully raised $205 million in a new funding round. The company develops specialized hardware designed to facilitate faster and more efficient communication between AI processors, a critical bottleneck in training large-scale machine learning models.

By optimizing the way data travels between chips, Cornelis intends to provide an alternative to the networking infrastructure currently utilized by Nvidia. As AI models grow in complexity, the demand for high-speed interconnects has surged, making networking technology a central battleground for tech companies looking to reduce reliance on single-vendor ecosystems. While the funding marks a significant milestone for the startup, the company faces the challenge of scaling its operations to compete with the established infrastructure dominance of industry leaders. The announcement highlights the ongoing trend of venture capital investment flowing into the 'picks and shovels' of the AI industry, specifically targeting hardware components that support the broader AI software ecosystem.