Embracing Local AI: Quadric’s Role in the Future of On-Device Inference

In an era where companies and governments are striving to reduce dependence on cloud infrastructure, the demand for tools that facilitate local AI operations is surging. Quadric, a promising chip-IP startup founded by veterans from the early bitcoin mining firm 21E6, is positioned at the forefront of this shift. By providing innovative on-device inference technology, Quadric aims to transition beyond its automotive roots into applications across laptops and industrial devices.

Transformative Growth in Licensing Revenue

With its unique approach, Quadric has witnessed remarkable financial growth. In 2025, the company reported licensing revenues between $15 million and $20 million, a significant increase from approximately $4 million in 2024. The company, headquartered in San Francisco and with a branch in Pune, India, projects revenues could rise to $35 million this year. This impressive financial trajectory has catapulted its post-money valuation to between $270 million and $300 million, up from about $100 million following its 2022 Series B funding round, as shared by CEO Veerbhan Kheterpal in a recent TechCrunch interview.

Attracting Investor Interest

This growth has not only boosted Quadric’s revenue but has also attracted significant investor interest. Recently, the startup announced a $30 million Series C funding round led by ACCELERATE Fund, managed by BEENEXT Capital Management. With this latest infusion of cash, Quadric’s total funding has reached $72 million. As industries, particularly chip manufacturers, seek solutions to migrate AI workloads away from centralized cloud infrastructures, Quadric stands as a beacon of hope for local server and device-based AI applications.

Expansion Beyond Automotive

Quadric’s journey began in the automotive sector, where its on-device AI technology enhances functionalities such as driver assistance. The explosion of transformer-based models in 2023 marked a significant turning point, prompting many businesses to reconsider their reliance on cloud-based technology. "Nvidia is a strong platform for data-center AI," Kheterpal notes. "We aimed to build a programmable infrastructure for on-device AI akin to a CUDA-like framework."

Unlike Nvidia, however, Quadric does not manufacture chips. Instead, it licenses its programmable AI processor intellectual property (IP), enabling customers to embed its technology directly into their silicon products. This approach is complemented by a robust software stack and toolchain, allowing various models—including vision and voice—to run locally.

More Read

Huawei’s Latest Ascend Chips: Revolutionizing the World’s Most Powerful Computing Cluster

Anthropic Forecasts $70 Billion Revenue by 2028, According to New Report

UK Utilizes AI Technology to Strengthen Arctic Security in Response to Increasing Threats

OpenAI Partners with UK Government to Explore Applications for AI Models

Half of xAI’s Founding Team Departs: Key Changes Within the Company

Diverse Customer Base

Quadric’s extensive customer base spans various industries, including printers, automotive, and AI-enabled laptops, with renowned companies like Kyocera and Japan’s Denso among its early adopters. The first products featuring Quadric’s technology are anticipated to launch this year, starting with laptops, as highlighted by Kheterpal.

Exploring Sovereign AI Strategies

As Quadric continues its expansion, it is exploring opportunities in the realm of "sovereign AI." This strategy aims to lessen dependence on U.S.-based infrastructure, particularly appealing to countries like India and Malaysia. The startup is guided by Moglix CEO Rahul Garg, who is a strategic investor, helping shape its sovereign approach in India. With nearly 70 employees globally, including about 40 in the U.S. and 10 in India, Quadric is poised for further growth.

Driving Distributed AI Adoption

The rising costs of centralized AI infrastructure have led many organizations to consider "distributed AI" solutions, where inference can be processed locally on devices or small on-premise servers. Kheterpal emphasizes this shift, noting the challenges faced by various countries in establishing hyperscale data centers. The World Economic Forum and EY have highlighted this movement towards local AI capabilities, further validating Quadric’s market approach.

Challenges and Adaptability in AI Hardware

Despite the potential, chip manufacturers face a significant hurdle: the pace at which AI models are evolving often outstrips hardware development cycles. Kheterpal argues that the need for programmable processor IP is pivotal, allowing customers to stay agile via software updates instead of incurring costs for hardware redesigns as model architectures shift.

Positioning Itself in a Competitive Landscape

In the competitive landscape dominated by companies like Qualcomm, Quadric offers a fresh alternative. Unlike Qualcomm, which typically integrates its AI technology into its processors, Quadric provides a versatile solution without locking customers into a specific hardware ecosystem. The startup’s programmable model allows clients to easily adapt to new AI architectures, maintaining relevance in an industry where timely innovation is crucial.

Steps Towards Realizing Long-term Potential

While Quadric has secured notable clients and agreements, its long-term success hinges on converting these partnerships into high-volume shipments and recurring royalties. Their future growth strategy is designed not just to compete, but to lead in the rapidly evolving landscape of on-device AI. By championing a local-first approach, Quadric is well-equipped to leverage the increasing demand for adaptable AI solutions, solidifying its standing in the tech ecosystem.

With the dynamic nature of AI technology, Quadric is navigating a path that emphasizes flexibility and innovation, carving out a niche that promises sustained relevance and growth in the market.

Inspired by: Source