Trending
Model-agnostic PII detection with LLMs Reduce inference cold starts on Amazon SageMaker HyperPod with model caching Instagram now lets users add tagged posts to their profile grid Build an end-to-end RFI questionnaire workflow using Amazon Quick Automate HelmGuard Raises $7.3M Seed Round | Forus Raises $150M at a $3B Valuation The Extinction Risk Preference Cascade: Quotes Lifesaving Lincoln Laboratory device wins 2026 Excellence in Technology Transfer Award DeepSeek launches V4.1-Flash with 1M-token context The Internet of Bodies is Coming – Your Body Already Knows – Life Sciences Today Podcast Episode 78 Nostrum’s data center in Badajoz gets the green light for electrical connection Reduce LLM latency with prefix-aware routing on Amazon SageMaker Inference Meta’s new AI app Muse tops 83,000 iOS downloads in the US Adobe beats revenue estimates but weak outlook hits shares Jacob Coxon Warns of Human Extinction and Triggers a Preference Cascade Anthropic report says AI model solved CAPTCHAs to upload malware

Nvidia AI interconnect darling Point2 bags $136m funding from Arm, others

AI data center interconnect specialist Point2 Technology has bagged Series B financing to the tune of $136 million.

Led by LB Investment, the funding round saw previous investors such as Nvidia, Molex, and Bosch Ventures joined by Arm and some unnamed participants.

money dollars

– Getty Images

Point2 plans to use the new funding to grow its engineering, operations, and commercial teams while accelerating the roll out of its active radio frequency (RF) cable (ARC), near‑packaged e-Tube (NPE), and co‑packaged e‑Tube (CPE) technologies, which are designed to power the next generation of rack-scale AI compute infrastructure.

Founded in 2016 and headquartered in San Jose, California, Point2 develops an RF signaling architecture that transmits data over plastic waveguides, as coupled with mixed‑signal interconnect system on chips (SoCs). The firm currently boasts partners such as Keysight Technologies and Taiwanese giant Foxconn.

Point2’s e‑Tube platform is presented as a connectivity solution to failing lasers and reportedly offers ten times the reach and two times lower cable volume than copper at a comparable cost, at three times lower power and lower cost, with a 1,000-fold lower latency than optics.

The offering may provide alternative connectivity amongst an ongoing laser squeeze fueled by AI-hungry hyperscaler demand. McKinsey claims production of 800 Gbps (800G) transceivers is projected to run up to 60 percent below demand until 2027, while shortfalls of up to 40 percent are likely for 1.6 Tbps (1.6T) devices through 2029.

“As AI systems scale and bandwidth demands reach terabit-per-second speeds, interconnect has become the defining bottleneck,” Point2 Technology CEO Sean Park noted on the Series B. “The latest investment … validates that our e‑Tube platform is emerging as a critical interconnect solution for next-generation AI data factories.”

“The next generation of AI infrastructure will rely on advances across the entire technology stack, from compute to memory and interconnect,” Paul Williamson, SVP for strategic ventures at Arm, added. “Point2 is addressing an increasingly important challenge as AI systems scale, and we are pleased to support their work in advancing the technologies that will underpin the future of AI infrastructure.”

 

Join the conversation

Your email address will not be published. Required fields are marked *