Silicon Showcase⁠™

Your shortlist

Nothing saved yet.

Save a solution from IP Finder, a solution listing, a solution page, or a provider’s solutions to compare it here.

Saved in this browser. Information is based on published sources. Confirm suitability and configuration with the provider.

Public source

Expedera Origin NPU / AI Accelerator IP

The Products Overview page reviewed on 12 November 2026 describes Origin Evolution NPU IP managing AI workloads from edge to automotive and data centers, customizable for PPA, supporting LLMs, CNNs, RNNs, and other models without hardware-specific model modifications.

View Official Source (opens in a new tab)Find IP for your application

Overview

Expedera Origin Evolution NPU IP is licensable AI accelerator IP. The official products overview states packet-based NPU cores scalable from edge devices to data centers with silicon-proven efficiency claims.

Stated architecture metrics include single-core performance up to 128 TFLOPS, 70–90% utilization on typical workloads (Llama, YOLO) measured on silicon, and 18 TOPS/W power efficiency, with ASIL-B readiness certification noted. Product line includes Origin E1 (~1 TOPS / LittleNPU), E2 (1–20 TOPS), E6 (16–32 TOPS), E8 (32–128 TOPS), Origin Evolution for Automotive (to 96 TFLOPS single-core), and TimbreAI T3 ultra-low-power inference for headsets. A TVM/Relax-based software stack supporting HuggingFace, Llama.cpp, PyTorch, ONNX, TensorFlow, and related frameworks is described. Official overview presents Origin NPU IP family for SoC licensing.

Technical capabilities

  • Published as design IP

    Licensable semiconductor IP. This listing is not a finished controller chip.

  • Finder function

    NPU / AI accelerator

  • Recorded aliases

    Origin Evolution; Origin E8; Expedera NPU

Lifecycle and categories

Solution categories
NPU IP
Supported versions
Origin E1/E2/E6/E8 / Evolution Automotive (TOPS ranges on page)
Compatibility
Full TVM/Relax software stack; edge to datacenter scaling. Source

Related solutions

AONDevices AON1000 Edge AI IP

NPU IP

AONDevices AON1000 is licensable edge-AI IP for always-on voice/sound sensing. The official site lists AON1000/AON1010/AON1020/AONDenoise under Products IP alongside chip and platform offerings.

View solution

Arm Ethos-U85

NPU IP

Licensable design IP. Embedded NPU for inference, including transformer workloads. It is not a CPU.

View solution

BrainChip Akida Neuromorphic Processor IP

NPU IP

BrainChip Akida is licensable neuromorphic neural processor IP. The official IP page describes Akida 1, Akida Pico, and Akida 2 variants with event-based acceleration and IP brochure downloads.

View solution

CEVA-NeuPro-Nano

NPU IP

CEVA-NeuPro-Nano is edge NPU processor IP, not a discrete accelerator chip and not a full SoC.

View solution

Cambricon-1H processor IP family

NPU IP

A published family of embedded AI processor IP for integration into chips.

View solution

EdgeCortix Dynamic Neural Accelerator II IP

NPU IP

EdgeCortix DNA-II is licensable neural network IP for ASIC, FPGA or SoC. The official DNA page describes run-time reconfigurable interconnects, scalable performance from 1K MACs, and MERA software co-optimization.

View solution

Related events

Explore promotion options