Point2 Technology Reaches $136M Series B for AI Interconnects

Point2 Technology has completed an additional Series B extension that brings its total Series B financing to $136M. LB Investment led the new financing, Arm joined as a new strategic investor, and Maverick Silicon continued its support.

The financing matters because AI infrastructure is running into a systems problem, not simply a chip problem. More accelerators per rack create heavier demands on the links moving data between them, and those links affect power, latency, reach, routing, reliability, and cost.

Point2 is building an alternative to conventional short-reach copper and optical interconnects. Its e-Tube platform sends radio-frequency signals through plastic waveguides, with Active RF Cable, near-packaged, and co-packaged form factors designed for future rack-scale compute systems.

What Point2 Announced

The August 10 financing adds another extension to a Series B that has grown in stages. Point2 announced an initial $22M Series B in May 2022, then disclosed an April 2026 extension led by Maverick Silicon, with NVentures and UMC Capital participating, that brought the Series B total to $76M.

The latest extension increases that total to $136M. Point2 did not disclose the incremental amount associated with this closing, the investors' individual check sizes, or a valuation, so the clean description is $136M in total Series B financing rather than a guessed standalone round amount.

The current investor roster is notable for its position across the infrastructure stack. Point2 cites NVIDIA, Arm, UMC Capital, Molex, Bosch Ventures, Maverick Silicon, LB Investment, and other new and existing investors, bringing together compute IP, semiconductors, connectivity manufacturing, and venture capital around the same interconnect thesis.

How e-Tube Fits the AI Infrastructure Stack

AI systems do not scale through accelerators alone. Data has to move between chips, boards, trays, and racks quickly enough that expensive compute is not waiting on the interconnect, and that movement has to stay inside increasingly unforgiving power and thermal budgets.

Point2's e-Tube architecture uses RF signaling over plastic waveguides. The company positions it between copper, which faces reach, weight, and routing constraints as data rates rise, and optical systems, which can add power, cost, latency, conversion, and reliability complexity.

Point2 says e-Tube offers 10X the reach of copper, 5X lower weight, and 2X lower cable volume at comparable cost. Against optics, the company reports 3X lower power, 3X lower cost, and 1000X lower latency without laser-related reliability concerns. Those figures are company-reported product comparisons, not independent production benchmarks, and broad commercialization still has to prove how the claims translate across customer architectures.

Why the Investor Mix Matters

Strategic capital does not guarantee adoption, but it can reveal where industry participants expect friction to move. Arm's participation connects Point2 to a compute architecture company, while prior backing from NVIDIA, UMC Capital, Molex, and Bosch Ventures links the company to accelerators, semiconductor manufacturing, interconnect products, and industrial systems.

That mix supports a larger shift in AI infrastructure spending. As raw compute capacity expands, more value can move toward the constraints surrounding it, including power delivery, cooling, memory, networking, and physical interconnect. A faster accelerator does not improve system economics if data movement forces the rack to sacrifice reach, density, energy efficiency, or reliability.

Point2 also enters this phase with third-party recognition. BloombergNEF selected the company as a 2026 Pioneer in the sustainable, scalable data-center-infrastructure category, describing its RF semiconductor interconnect technology as a path toward faster and more energy-efficient AI-data-center communication.

What the Funding Changes

Point2 plans to use the capital to expand its engineering, systems, operations, and business teams. The company is also accelerating commercialization of three related product paths: Active RF Cable, near-packaged e-Tube, and co-packaged e-Tube solutions for rack-scale compute systems.

That work moves the company from a technology argument toward an execution test. Cable and interconnect products have to survive qualification, manufacturing, packaging, systems integration, and customer deployment requirements, and strategic investors can help open technical conversations without removing any of those gates.

The current leadership team reflects that transition. Co-founder and CEO Sean Park leads the company, CTO Thomas Cho brings decades of RF, analog, mixed-signal, and high-speed communications experience, and founder and chairman HM Bae connects the business to its KAIST research roots.

What This Signals for AI Data Centers

The headline is a $136M Series B total, but the more durable story is where infrastructure investors are placing their attention. AI data centers are becoming integrated systems in which the economic value of compute depends on memory, networking, power, cooling, and the physical links connecting every component.

Point2 is betting that RF-over-plastic-waveguide interconnects can become one of those foundational links. The financing gives the company more resources and strategic access to make that case, but the next evidence will come from qualification, manufacturability, customer adoption, and real deployment performance.

For operators and investors, that is the right frame. The round does not settle whether e-Tube becomes a rack-scale standard, but it shows that a group of sophisticated semiconductor and infrastructure backers considers interconnect important enough to fund before the bottleneck becomes even more expensive.