IP Catalog > CNN Accelerator

CNN accelerator IP

Filter
Filter

Login required.

Sign in

Search (11)
Compare 11 IP from 8 vendors (1 - 10)
  • Convolutional Neural Network (CNN) Compact Accelerator
    • Support convolution layer, max pooling layer, batch normalization layer and full connect layer
    • Configurable bit width of weight (16 bit, 1 bit)
    Block Diagram -- Convolutional Neural Network (CNN) Compact Accelerator
  • AI Accelerator Specifically for CNN
    • A specialized hardware with controlled throughput and hardware cost/resources, utilizing parameterizeable layers, configurable weights, and precision settings to support fixed-point operations.
    • This hardware aim to accelerate inference operations, particulary for CNNs such as LeNet-5, VGG-16, VGG-19, AlexNet, ResNet-50, etc.
    Block Diagram -- AI Accelerator Specifically for CNN
  • High performance-efficient deep learning accelerator for edge and end-point inference
    • Configurable MACs from 32 to 4096 (INT8)
    • Maximum performance 8 TOPS at 1GHz
    • Configurable local memory: 16KB to 4MB
    Block Diagram -- High performance-efficient deep learning accelerator for edge and end-point inference
  • Accelerator for Convolutional Neural Networks
    • Include VGG, ResNet, MobileNet, and other custom use cases.
  • AI Accelerator
    • Independent of external controller
    • Accelerates high dimensional tensors
    • Highly parallel with multi-tasking or multiple data sources
    • Optimized for performance / power / area
  • Sensor Fusion IP
    • Kalman Filter
    • Extended Kalman Filter
    • CNN
  • Neural engine IP - Tiny and Mighty
    • The Origin E1 NPUs are individually customized to various neural networks commonly deployed in edge devices, including home appliances, smartphones, and security cameras.
    • For products like these that require dedicated AI processing that minimizes power consumption, silicon area, and system cost, E1 cores offer the lowest power consumption and area in a 1 TOPS engine.
    Block Diagram -- Neural engine IP - Tiny and Mighty
  • 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
  • Neural engine IP - The Cutting Edge in On-Device AI
    • The Origin E6 is a versatile NPU that is customized to match the needs of next-generation smartphones, automobiles, AV/VR, and consumer devices.
    • With support for video, audio, and text-based AI networks, including standard, custom, and proprietary networks, the E6 is the ideal hardware/software co-designed platform for chip architects and AI developers.
    • It offers broad native support for current and emerging AI models, and achieves ultra-efficient workload scheduling and memory management, with up to 90% processor utilization—avoiding dark silicon waste.
    Block Diagram -- Neural engine IP - The Cutting Edge in On-Device AI
  • Neural engine IP - Balanced Performance for AI Inference
    • The Origin™ E2 is a family of power and area optimized NPU IP cores designed for devices like smartphones and edge nodes.
    • It supports video—with resolutions up to 4K and beyond— audio, and text-based neural networks, including public, custom, and proprietary networks.

     

    Block Diagram -- Neural engine IP - Balanced Performance for AI Inference
×
Semiconductor IP