IP Catalog > Processor IP > NPU Processor IP Cores

NPU Processor IP Cores

NPU (Neural Processing Unit) Processor IP cores provide high-performance computing power for tasks such as image recognition, natural language processing, and data analysis, enabling real-time AI processing at the edge.

All offers in NPU Processor IP Cores
Filter
Filter

Login required.

Sign in

Search (44)
Compare 44 NPU Processor IP Cores from 10 vendors (1 - 10)
  • GPNPU Processor IP - 32 to 864TOPs
    • 32 to 864TOPs
    • (Dual, Quad, Octo Core) Up to 256K MACs
    • Hybrid Von Neuman + 2D SIMD matrix architecture
    • 64b Instruction word, single instruction issue per clock
    • 7-stage, in-order pipeline
    • Scalar / vector / matrix instructions modelessly intermixed with granular predication
    Block Diagram -- GPNPU Processor IP - 32 to 864TOPs
  • GPNPU Processor IP - 16 to 108 TOPs
    • 16 to 108 TOPs
    • 8K / 16K / 32K MACs plus 1024 ALUs
    Block Diagram -- GPNPU Processor IP - 16 to 108 TOPs
  • GPNPU Processor IP - 1 to 7 TOPs
    • 1 to 7TOPs
    • 512/ 1K/ 2K/ 8K MACs plus 64 ALUs
    Block Diagram -- GPNPU Processor IP - 1 to 7 TOPs
  • GPNPU Processor IP - 4 to 28 TOPs
    • 4 to 28 TOPs
    • 2K/ 4K/ 8K MACs plus 256 ALUs
    Block Diagram -- GPNPU Processor IP - 4 to 28 TOPs
  • NPU IP Core for Mobile
    • Origin Evolution™ for Mobile offers out-of-the-box compatibility with popular LLM and CNN networks. Attention-based processing optimization and advanced memory management ensure optimal AI performance across a variety of today’s standard and emerging neural networks.
    • Featuring a hardware and software co-designed architecture, Origin Evolution for Mobile scales to 64 TFLOPS in a single core.
    Block Diagram -- NPU IP Core for Mobile
  • Specialized Video Processing NPU IP
    • Highly optimized for CNN-based image processing application
    • Fully programmable processing core: Instruction level coding with Chips&Media proprietary Instruction Set Architecture (ISA)
    • 16-bit floating point arithmetic unit
    • Minimum bandwidth consumption
    Block Diagram -- Specialized Video Processing NPU IP
  • NPU IP Core for Edge
    • Origin Evolution™ for Edge offers out-of-the-box compatibility with today's most popular LLM and CNN networks. Attention-based processing optimization and advanced memory management ensure optimal AI performance across a variety of networks and representations.
    • Featuring a hardware and software co-designed architecture, Origin Evolution for Edge scales to 32 TFLOPS in a single core to address the most advanced edge inference needs.
    Block Diagram -- NPU IP Core for Edge
  • NPU IP Core for Data Center
    • Origin Evolution™ for Data Center offers out-of-the-box compatibility with popular LLM and CNN networks. Attention-based processing optimization and advanced memory management ensure optimal AI performance across a variety of today’s standard and emerging neural networks. Featuring a hardware and software co-designed architecture, Origin Evolution for Data Center scales to 128 TFLOPS in a single core, with multi-core performance to PetaFLOPs.
    Block Diagram -- NPU IP Core for Data Center
  • NPU IP Core for Automotive
    • Origin Evolution™ for Automotive offers out-of-the-box compatibility with popular LLM and CNN networks. Attention-based processing optimization and advanced memory management ensure optimal AI performance across a variety of today’s standard and emerging neural networks.
    • Featuring a hardware and software co-designed architecture, Origin Evolution for Automotive scales to 96 TFLOPS in a single core, with multi-core performance to PetaFLOPs.
    Block Diagram -- NPU IP Core for Automotive
  • Neural engine IP - AI Inference for the Highest Performing Systems
    • The Origin E8 is a family of NPU IP inference cores designed for the most performance-intensive applications, including automotive and data centers.
    • With its ability to run multiple networks concurrently with zero penalty context switching, the E8 excels when high performance, low latency, and efficient processor utilization are required.
    • Unlike other IPs that rely on tiling to scale performance—introducing associated power, memory sharing, and area penalties—the E8 offers single-core performance of up to 128 TOPS, delivering the computational capability required by the most advanced LLM and ADAS implementations.
    Block Diagram -- Neural engine IP - AI Inference for the Highest Performing Systems
×
Semiconductor IP