Mixx Technologies Raises $33M to Redefine Optical Connectivity for the AI Infrastructure Era

Deep-tech startup Mixx Technologies has raised $33 million in an oversubscribed Series A round, securing backing from some of the world’s leading semiconductor and AI infrastructure investors. The round was led by Singapore’s ICM HPQC Fund with participation from TDK Ventures, Systemiq Capital, Applied Ventures, Banpu Innovation & Ventures, G Vision Capital, Ajinomoto Group Ventures, AVITIC Innovation Fund and other strategic partners.

The investment reflects a growing consensus across the AI ecosystem: as models scale exponentially, data movement, not compute, has become the defining bottleneck of modern AI infrastructure. Mixx aims to solve that bottleneck with a new class of silicon-integrated optical connectivity that re-architects the way hyperscale data centers move information between trillions of interconnected nodes.

Founded by industry veterans behind Intel’s silicon photonics transceivers and Broadcom’s first co-packaged optics (CPO) platforms, Mixx Technologies is positioning itself as the foundational layer for next-generation AI systems that demand unprecedented speed, parallelism and energy efficiency.

The Connectivity Bottleneck Holding Back AI

AI infrastructure is entering a new stage of scale, driven by massive generative models, high-throughput inference workloads and increasingly interconnected compute clusters. With GPUs now producing data faster than electrical interconnects can transport it, hyperscale AI facilities are running into hard physical limits. According to Mixx CEO and co-founder Vivek Raghuraman, this is where today’s architectures fall apart.

“As AI infrastructure scales into the exabyte era, the very metrics of performance are shifting. What once centered on link speeds and component efficiency must now account for system-wide power, latency, and reliability at the data center level.”

Electrical interconnects: copper SerDes, pluggable modules and legacy switching are struggling to keep up with bandwidth demands and power constraints. Data movement now consumes up to 70% of total data center energy, according to industry estimates.

Mixx’s core thesis: the future of AI is bottlenecked by connectivity, not compute.

Mixx’s Breakthrough: Silicon-Integrated Optical Connectivity

At the center of Mixx’s platform is HBxIO™, a silicon-integrated optical engine engineered for ultra-high-radix, system-scale connectivity. Rather than treat photonics, packaging and system architecture as separate domains, Mixx integrates them into a unified platform designed to flatten network topologies, reduce energy consumption and dramatically improve throughput. Key capabilities include:

Switchless AI clusters: Mixx enables a flattened, high-radix architecture that reduces reliance on traditional network switches. The result:

• • 4× higher port counts compared to copper CPO
  • Up to 32× improvement in AI compute efficiency
  • Lower latency and reduced network complexity

3.5D Integration for ultra-short data paths: Optics are integrated directly onto ASICs, eliminating unnecessary components and failure points. This achieves:

• Up to 75% lower power consumption
• 2× lower latency
• Higher reliability at scale

A composable fabric for disaggregated AI systems: HBxIO™ allows compute, memory and accelerators to be interconnected dynamically, a critical architectural shift as inference workloads surpass training in scale and complexity.

Fully interoperable, open-standards design: Mixx is purpose-built to work within existing data center ecosystems, accelerating adoption and reducing integration overhead for cloud providers. Mixx describes this as “transforming the path from transistor to fiber”, a claim backed by its deep engineering pedigree.

A Founding Team With a Deep-Tech Track Record

Mixx was founded by the technologists behind several breakthroughs in silicon photonics and co-packaged optics, including Intel’s commercial silicon photonics transceivers and Broadcom’s first-generation CPO for network switches.

This collective experience has given the company a rare advantage: the ability to take optical connectivity from concept to scaled deployment in hyperscale environments.

Co-founder and Chief Product Officer Dr. Rebecca Schaevitz explains the mission: “We’re transforming the path from transistor to fiber with silicon-integrated optics, flattening the network and eliminating bottlenecks at the core. Our connectivity fabric is engineered to scale as aggressively as modern AI compute.“

Investors See Mixx as a Catalytic Technology

Investors across the deep-tech and AI infrastructure ecosystem express extraordinary confidence in Mixx’s approach.

Matthew Gould, Portfolio Manager at ICM HPQC Fund, said: “AI connectivity infrastructure is being redefined in real time and Mixx has brought together the most capable team to lead that transformation.”

TDK Ventures echoed this view: “Mixx’s pragmatic approach makes manufacturability seamless for the AI era, delivering energy efficiency and a multi-generation roadmap for performance improvement.”

Systemiq Capital emphasized the importance of sustainability, noting: “Mixx is delivering the right technology at exactly the right time, a system-level optical architecture designed to optimize power and economics for hyperscale AI.”

Scaling to Meet Exabyte-Scale AI Demand

With the new capital, Mixx will:

• Expand R&D centers in the US, India and Taiwan
• Advance product development for hyperscale customers
• Grow global engineering and manufacturing capacity
• Deepen partnerships with AI hardware and data center ecosystem players

As inference workloads expand and models scale beyond trillions of parameters, the industry is looking for infrastructure capable of supporting the next decade of AI growth. Mixx believes its silicon-integrated optical platform is that foundation.

Mixx Technologies stands out in a crowded AI hardware ecosystem by focusing on the problem others have long avoided: moving data efficiently at scale. While many deep-tech startups chase incremental performance gains in compute, Mixx is addressing the systemic bottleneck that determines whether large-scale AI is even viable.

If Mixx delivers on its promises, particularly switchless clusters and 3.5D optical integration, it could become one of the most important infrastructure companies of the next decade. However, achieving manufacturable, reliable and cost-effective silicon-integrated optics at scale remains one of the hardest engineering challenges in the industry. The opportunity is massive, but so are the hurdles. Still, this Series A round signals strong belief from investors who deeply understand the space. Mixx is building the connective tissue for the next generation of AI systems.