See the 2025 Best Edge AI Processor IP at the Embedded Vision Summit

Burlingame, CA – May 12, 2025 – Quadric® today announced that it will showcase its Chimera^TM general-purpose neural processing unit (GPNPU) at the Embedded Vision Summit, May 20-22, 2025, at the Santa Clara Convention Center.

The Chimera GPNPU was recently named the 2025 Best Edge AI Processor IP by the Edge AI and Vision Alliance. Quadric will be in booth #821 at the exhibition, which is open Wednesday, May 21, from 12:30 – 7:30 pm and Thursday, May 22, from 11 am – 5 pm.

A highlight at the event will be the talk by Nigel Drego, Co-Founder and Chief Technology Officer at Quadric. “ONNX and Python to C++: State-of-the-Art Graph Compilation” on Thursday May 22 at 12:00 noon. Mr. Drego will discuss how the Chimera Graph Compiler uniquely compiles ONNX graphs into C++ representations targeting the Chimera GPNPU and how the entire Chimera Software Development Kit (SDK) combines those ONNX graphs plus DSP C++ code and Python code into one merged, optimized code stream. The Chimera tools deliver far greater programmer productivity and superior results when compared to competing AI architectures that employ two or three incompatible processing elements with separate, disjoint CPU, DSP and NPU toolchains.

Quadric’s Chimera processor is the only fully C++ programmable AI/ML inference solution that delivers the energy efficiency of a dedicated neural network accelerator plus the full flexibility of a programmable solution in a single architecture. A significant advantage of Quadric’s solution is that neural network graphs and C++ code are merged into a single, fully programmable software application. Only one tool chain is required for scalar, vector, and matrix computations.

About Quadric

Quadric Inc. is the leading licensor of fully programmable, general-purpose neural processor IP (GPNPU) capable of running both machine learning inference workloads and traditional DSP and control algorithms. Quadric’s unified hardware and software architecture is optimized for on-device ML inference, delivering up to 864 TOPs in both commercial grade and automotive-grade safety-enhanced versions. Learn more at: www.quadric.io.