Companies and governments are increasingly seeking solutions to run artificial intelligence (AI) workloads locally, aiming to significantly reduce burgeoning cloud infrastructure costs and bolster sovereign technological capabilities. Leading this transformative shift is Quadric, a chip-IP startup established by veterans from the early bitcoin mining firm 21E6. Quadric is strategically scaling its on-device inference technology beyond its initial automotive focus, extending its reach into a diverse range of sectors including laptops and industrial devices.

This strategic expansion is already yielding substantial results for Quadric. The company reported impressive licensing revenue between $15 million and $20 million in 2025, marking a significant increase from approximately $4 million in 2024. CEO Veerbhan Kheterpal shared that Quadric is targeting an ambitious $35 million in revenue this year, as it meticulously builds a royalty-driven on-device AI business model. This robust growth has elevated the company's post-money valuation to an estimated $270 million to $300 million, a substantial leap from its Series B valuation of around $100 million in 2022. Quadric's financial momentum recently culminated in a successful $30 million Series C funding round, led by ACCELERATE Fund, managed by BEENEXT Capital Management, bringing its total funding to $72 million.

The impetus behind this investment and Quadric's success stems from a broader industry trend where investors and chipmakers are actively seeking methods to decentralize AI workloads, moving them away from centralized cloud infrastructure onto individual devices and local servers. Kheterpal highlighted that while Nvidia remains a dominant platform for data-center AI, Quadric is focused on developing a similar programmable, 'CUDA-like' infrastructure specifically tailored for on-device AI. Unlike Nvidia, Quadric does not manufacture chips directly. Instead, it licenses its programmable AI processor IP – essentially a blueprint – which customers can integrate into their own silicon. This IP is complemented by a comprehensive software stack and toolchain, enabling the efficient execution of various AI models, including vision and voice, directly on the device. Quadric’s technology is notably chip-agnostic, offering broad compatibility.

Quadric's journey began in the automotive sector, where on-device AI is crucial for real-time functions such as advanced driver assistance systems. However, the proliferation of transformer-based models in 2023 spurred a significant business inflection, pushing the demand for on-device inference into virtually 'everything' over the past 18 months. The startup now boasts a diverse customer base spanning printers, cars, and AI laptops, with notable clients including Kyocera and Japan's auto supplier Denso, which produces chips for Toyota vehicles. The first products featuring Quadric's technology, starting with laptops, are anticipated to ship later this year.

Beyond traditional commercial deployments, Quadric is also actively exploring markets engaged in 'sovereign AI' strategies. These initiatives aim to reduce reliance on foreign, particularly U.S.-based, AI infrastructure by fostering domestic AI capabilities. Kheterpal confirmed the company's exploration of opportunities in countries like India and Malaysia, with Moglix CEO Rahul Garg acting as a strategic investor guiding its approach in India. The shift towards distributed AI is primarily driven by the escalating costs associated with centralized AI infrastructure and the considerable challenges many nations face in establishing hyperscale data centers. Reports from the World Economic Forum and EY corroborate this growing trend, emphasizing the movement of AI inference closer to users and away from purely centralized architectures.

A critical challenge for contemporary chipmakers is the rapid evolution of AI models, which often outpaces the lengthy hardware design cycles. Kheterpal asserts that customers require programmable processor IP that can adapt to these rapid changes through software updates, rather than necessitating costly hardware redesigns every time AI architectures shift. Quadric positions itself as a distinct alternative to established chip vendors like Qualcomm, whose integrated AI technology often locks customers into their proprietary silicon, and IP suppliers such as Synopsys and Cadence, whose neural processing engine blocks are frequently perceived as difficult to program. Quadric’s programmable methodology grants customers the crucial ability to support new AI models via software updates, providing a significant competitive edge in an industry where chip development can take years, while model architectures can transform in a matter of months.

Despite its significant progress and strategic advantages, Quadric remains in the earlier stages of its buildout. The company currently has a handful of signed customers, and its longer-term growth trajectory and upside are contingent upon successfully converting existing licensing agreements into high-volume product shipments and consistent recurring royalties. This foundational period is critical for Quadric to solidify its market position and realize its full potential in the burgeoning on-device AI landscape.