NPU IP Cores

Sign in required. Page 2 and beyond are available to signed-in users.

NPU (Neural Processing Unit) IP cores are specialized AI processors designed to accelerate machine learning inference in modern SoC and ASIC designs.

These IP cores implement highly optimized architectures for tensor operations, including matrix multiplication and convolution, enabling efficient execution of deep learning models.

This catalog allows you to compare NPU IP cores from leading vendors based on TOPS performance, energy efficiency (TOPS/W), latency, and supported neural network frameworks.

Whether you are designing AI-enabled SoCs, edge devices, smart cameras, or automotive systems, you can find the right NPU IP for your AI acceleration needs.

 
All offers in NPU
Clear

Compare 62 NPU IP Cores from 25 vendors

What is an NPU (Neural Processing Unit)?

An NPU is a dedicated hardware accelerator optimized for executing neural network workloads within SoCs and ASICs.

NPUs are designed to efficiently process large-scale parallel computations required by deep learning models, delivering high performance at low power.

Main functions of NPU IP

Tensor and matrix processing

NPU IP accelerates matrix multiplications and tensor operations, which are the core computations in neural networks.

Deep learning inference

NPUs execute trained models for tasks such as image recognition, speech processing, and data classification.

Low-precision computation

NPU IP supports quantized data formats (INT8, FP16, etc.), improving performance and energy efficiency.

Parallel processing architecture

Highly parallel compute units enable efficient execution of large neural network models.

Key features of NPU IP cores

  • High AI performance (TOPS)
  • Excellent energy efficiency (TOPS/W)
  • Support for multiple neural network models
  • Low-latency inference
  • Integration with CPU, GPU, and memory subsystems

How to choose NPU IP

Selecting the right NPU IP depends on your AI workload:

  • Performance requirements (TOPS)
  • Energy efficiency targets
  • Supported data formats (INT8, FP16, etc.)
  • Latency constraints
  • Software ecosystem and framework support

Applications of NPU IP cores

  • Computer vision systems
  • Smart cameras and surveillance
  • Voice and speech processing
  • Automotive AI and ADAS
  • IoT and edge AI devices

Compare NPU IP cores from leading vendors

This catalog provides access to a wide range of NPU IP cores from leading semiconductor IP providers.

Use filters to compare solutions based on performance, power efficiency, latency, and integration requirements.

Frequently asked questions about NPU IP cores

What is an NPU?

A Neural Processing Unit is a processor optimized for executing machine learning and deep learning workloads.

What is the difference between NPU and GPU?

NPUs are specialized for AI workloads with higher efficiency, while GPUs are more general-purpose parallel processors.

Where are NPUs used?

NPUs are used in AI-enabled devices such as smart cameras, smartphones, automotive systems, and IoT devices.

What performance metrics are important?

Key metrics include TOPS, TOPS/W, latency, and supported data precision formats.

How do I choose an NPU IP?

You should consider performance, power efficiency, supported models, latency, and software ecosystem.

×
Semiconductor IP