SpaceX Veterans Launch $50M Startup to Revolutionize AI Data Center Optical Links

In a significant development at the intersection of aerospace engineering and artificial intelligence infrastructure, a new startup named Mesh has raised $50 million in Series A funding led by prominent venture capital firms, including Sequoia Capital and a16z. The company, founded by a team of veteran engineers from SpaceX, is pioneering the mass production of next-generation optical transceivers designed to address the escalating bandwidth demands of AI data centers.

According to TechCrunch, the founding team includes former SpaceX leads from the Starlink satellite communications and propulsion systems divisions, who bring deep expertise in high-reliability photonics, thermal management, and ultra-precision manufacturing. These skills, honed in the extreme environments of spaceflight, are now being repurposed to solve one of the most pressing challenges in AI computing: the inability of conventional copper-based interconnects to keep pace with the exponential growth of AI model training workloads.

Mesh’s proprietary optical transceivers utilize a novel silicon photonics architecture that reduces power consumption by up to 40% compared to current industry standards while increasing data throughput to 1.6 terabits per second per port. The company’s innovation lies not only in the optical design but in its scalable manufacturing process — adapted from SpaceX’s rapid iteration and vertical integration model for rocket components. This allows Mesh to produce transceivers at a fraction of the cost and time of traditional semiconductor firms, enabling rapid deployment across hyperscale data centers.

"We didn’t build rockets to go to Mars; we built systems that work under extreme constraints," said Elena Vasquez, CEO and former Starlink hardware lead at SpaceX. "The same principles apply in AI data centers — latency is the new gravity. If you can’t move data fast enough between GPUs, the entire system stalls. We’re bringing space-grade reliability and efficiency to the heart of the AI revolution."

The global demand for such technology is surging. Analysts from Gartner estimate that by 2028, over 70% of new AI data centers will require optical interconnects exceeding 800 Gbps per link — far beyond what today’s commercial transceivers can reliably deliver at scale. Current suppliers like Broadcom and Intel are struggling to meet this demand due to complex supply chains and long lead times.

Mesh’s initial customers include major cloud providers and AI research labs in North America and Europe. The company has also entered preliminary discussions with Middle Eastern sovereign wealth funds, including those in Saudi Arabia and the UAE, which are investing heavily in AI infrastructure as part of their economic diversification strategies. According to a report from PressBee, Mesh is exploring a regional manufacturing hub in the Kingdom of Saudi Arabia to serve the growing Middle Eastern AI market.

The $50 million funding round will be used to scale production facilities in Austin, Texas, and to establish a dedicated R&D center focused on co-packaged optics and AI-optimized network topologies. The company has already filed 12 patents covering its photonic chip design, packaging techniques, and automated calibration systems.

Industry observers note that Mesh represents a growing trend of aerospace engineers transitioning into the AI hardware sector — a movement that could redefine the future of computing infrastructure. With global AI spending projected to exceed $1 trillion by 2030, companies that can deliver faster, cooler, and cheaper optical links will become indispensable.

Mash’s success could signal the dawn of a new era — one where technologies developed for the final frontier are now powering the next revolution on Earth.