Participated · Series A · Jul 2026

Founded in 2024 and headquartered in Santa Clara, ChipAgents develops an AI-native platform that applies domain-specific autonomous agents to semiconductor design and verification. Its agents transform specifications and code into production-ready RTL and verification assets and perform automated root-cause analysis, claiming to reduce design cycles by over 50% and compress tasks that once took months into days or hours. The company reported rapid commercial adoption, with deployments at more than 120 semiconductor companies including MediaTek and Micron and 6x ARR growth in the first half of 2026. ChipAgents positions its technology as moving beyond AI copilots to agents that can independently plan and execute multi-step engineering workflows. The company plans to use recent funding to scale customer deployments, expand engineering and go-to-market headcount, and accelerate product development to meet rising global demand.

Participated · Series A · Jul 2026

OXMIQ develops a licensable GPU core, OxCore™, that integrates a CUDA-compatible GPU engine, a tensor processing engine, and an orchestration engine to enable configurable AI silicon. Its OxQuilt™ chiplet integration architecture supports heterogeneous compute and memory configurations across foundries and packaging options, and OxPython™ and OxCapsule™ provide software portability and orchestration for running existing CUDA and PyTorch code. OxCore is running on FPGA today with live demonstrations available, and the company positions an IP-first model to generate revenue from customer engagements while conserving capital. OXMIQ is working with semiconductor companies, neoclouds, AI system builders, and partners such as MediaTek and AM Intelligence Labs on initiatives including a renewable-powered AI factory collaboration. The company has strengthened its governance with additions to its board and advisory ranks to support customer integration and scaling. Financially, OXMIQ has raised $60 million in total capital following the $35 million Series A.

Participated · Series B · Jun 2026

Stathera develops MEMS-based silicon timing products built on its proprietary DualMode® architecture, aiming to replace legacy quartz timing in a range of electronics from smartphones and wearables to AI data centers. Its GEN2 32.768 kHz product line is moving into mass production and broader GEN2 portfolio products are sampling with Tier 1 OEMs. Stathera's silicon oscillators are manufactured in standard semiconductor fabs and are positioned to offer smaller footprint, higher reliability, improved shock and vibration resilience, and no external load capacitors compared with quartz. The company is launching a GEN3 program purpose-built for the performance and reliability demands of AI-CED applications and is targeting first customer samples in 2028. Stathera positions the AI data center timing opportunity as a fast-growing segment within an estimated US$11 billion annual timing market and cites an industry estimate that AI data center time-synchronization could represent a cumulative US$1.5 billion opportunity by 2030. The company is headquartered in Montreal and has raised US$75 million in total funding following the Series B.

Led · Series C · Jun 2026

HyperLight develops high-performance integrated photonics based on thin-film lithium niobate (TFLN), commercialized as its TFLN Chiplet™ Platform. The platform targets a range of optical interconnect applications including data center pluggables, coherent datacom and telecom modules, and co-packaged optics, and offers products supporting 200G-per-lane operation with 400G-per-lane solutions now sampling. The company emphasizes ultra-high modulation bandwidth, CMOS-level drive voltage, and ultra-low optical loss to reduce power consumption at higher lane speeds. HyperLight has established a strategic manufacturing partnership with UMC and Wavetek for high-volume foundry production on both 6-inch and 8-inch wafers. With the $80 million Series C, the company plans to expand manufacturing capacity, accelerate customer qualification, scale its platform, and deepen partnerships across foundry, semiconductor, networking, and systems integration domains.

Participated · Seed · May 2026

RadixArk is a spin-out from UC Berkeley that maintains SGLang, an open-source engine designed to make the inference phase of large language models markedly faster and cheaper. The platform is already in use by teams at xAI and Cursor to boost training and real-time inference performance. Building on this foundation, RadixArk will keep SGLang free while offering paid cloud hosting and releasing new products such as Miles, a reinforcement-learning framework that helps models improve over time. The effort began in 2023 under Databricks co-founder Ion Stoica and has now transitioned into an independent commercial entity. Although much of its software remains open source, the company has introduced revenue-generating hosting services to capitalize on enterprise demand for AI infrastructure. RadixArk’s latest fundraising values the firm at roughly $400 million, giving it resources to expand product development and go-to-market efforts.