top of page
Mobilint qb
Full-stack SDK Support
We are proud to present qb, our tailor-made Software Development Kit (SDK). Our main goal was to design a comprehensive pack to support our clients’ NPU application and development experience to be intuitive, quick, and accessible. Through strenuous trials and evaluations at MLPerf™, qb has demonstrated its many strengths. It is the verified SDK to meet all your comprehensive needs. Empower yourself instantaneously for an emerging AI world by implementing a deep learning application with NPU acceleration accompanied by qb.
.png)
Access 300+ models
Supported Frameworks
Mobilint Compiler Suite
Mobilint Compiler Suite's proprietary techniques optimize the original model while retaining its accuracy.
Mobilint Executable(MXQ)
An NPU-ready format for Mobilint runtime.
Multiple Instances
Mobilint NPUs can be plugged into the host device in multiple instances depending on the application.
NPU acceleration accompanied by qb
Key benefits
Mobilint qb simplifies deployment for engineers and optimizes performance for end users.
01
Comprehensive Framework Support
Seamlessly integrates with PyTorch, TensorFlow, TFLite, ONNX, Keras, etc.
02
Advanced Quantization Technique
Maintains over 99 % of the original model performance even after Int8 quantization.
03
Cross-Platform Compatibility
Supports both Windows and Ubuntu for flexible deployment options.
04
User-Friendly API
Intuitive API for quick NPU integration, requiring minimal coding effort.
05
Extensive Model Support
Supports over 300 models, suitable for various domains like detection, segmentation, classification, and super resolution.
bottom of page