intel/npu-nn-cost-model
Library for modelling performance costs of different Neural Network workloads on NPU devices
This project helps embedded systems engineers and AI solution architects predict the performance of different neural network workloads on Intel's Neural Processing Units (NPUs) before deploying them. By providing details about a neural network operation (like input/output dimensions, kernel size, and execution mode), it outputs the estimated DPU cycles, helping users optimize their AI models for NPU devices. This is for professionals who design and optimize AI systems for specialized hardware.
Use this if you are developing or deploying neural network models on Intel NPU devices and need to accurately estimate their performance costs to make informed design choices.
Not ideal if you are a data scientist primarily focused on model training and experimentation without a direct need to optimize for specific edge AI hardware performance characteristics.
Stars
34
Forks
4
Language
C++
License
Apache-2.0
Category
Last pushed
Mar 11, 2026
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/ml-frameworks/intel/npu-nn-cost-model"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Higher-rated alternatives
microsoft/onnxruntime
ONNX Runtime: cross-platform, high performance ML inferencing and training accelerator
onnx/onnx
Open standard for machine learning interoperability
PINTO0309/onnx2tf
Self-Created Tools to convert ONNX files (NCHW) to TensorFlow/TFLite/Keras format (NHWC). The...
NVIDIA/TensorRT
NVIDIA® TensorRT™ is an SDK for high-performance deep learning inference on NVIDIA GPUs. This...
onnx/onnxmltools
ONNXMLTools enables conversion of models to ONNX