Quadric has developed the Chimera GPNPU, a licensable processor architecture that integrates on-device machine learning inference with the ability to run complex C++ code without requiring code partitioning across multiple processor types. This technology scales from 1 to 864 TOPs and supports all machine learning models, including classical networks and large language models, streamlining SoC design and accelerating model porting.
Funding
$48.3M raised to dateRaised to date based on public sources. This may differ from the amount the company actually raised and is based only on what is publicly available on the internet.
Founders
Product
Problem
Modern System on Chip (SoC) designs for AI inference often require partitioning code across multiple processor types, increasing complexity and development time. Existing solutions struggle to efficiently handle both machine learning inference and complex C++ code on a single architecture, leading to performance bottlenecks and integration challenges.
Solution
Quadric offers the Chimera GPNPU, a licensable processor architecture designed to streamline SoC design by integrating on-device machine learning inference with the ability to run complex C++ code without code partitioning. The Chimera GPNPU scales from 1 to 864 TOPs and supports a wide range of machine learning models, including classical networks, vision transformers, and large language models (LLMs). This unified architecture simplifies application development and accelerates the porting of new ML models, reducing the need for multiple specialized processors.
Target Audience
The primary target audience includes SoC designers and system architects in markets such as automotive, edge computing, and embedded systems who require high-performance, flexible, and scalable AI inference capabilities.
Features
- Single, unified architecture for both ML inference and complex C++ code execution
- Scalable performance from 1 to 864 TOPs
- Support for all machine learning models, including classical networks, vision transformers, and LLMs
- Chimera DevStudio for AI software simulation and SoC design visualization
- Safety-enhanced versions available for automotive applications (ASIL-ready cores)
- Simplified SoC hardware design and accelerated ML model porting