An Open-Source ML-Based Full-Stack Optimization Framework for Machine Learning Accelerators

Keyphrases

• Optimization Framework 100%
• Machine Learning Based 100%
• System Metrics 100%
• Stack Optimization 100%
• Machine Learning Accelerator 100%
• Machine Learning 66%
• Deep Neural Network 66%
• Power Performance 66%
• Design Space Exploration 66%
• Area Metric 66%
• Fully Automated 33%
• Unified Approach 33%
• Physical Design 33%
• Design Basis 33%
• Machine Learning Algorithms 33%
• Area Analysis 33%
• Learning-based 33%
• Prediction Error 33%
• Performance Simulation 33%
• Automated Search 33%
• Exploration Method 33%
• ASIC Implementation 33%
• Automated Design 33%
• Network Machine Learning 33%
• Power Performance Analysis 33%
• Prediction Framework 33%
• Hardware Accelerator 33%
• Deep Learning Accelerator 33%

Computer Science

• Open Source 100%
• Machine Learning 100%
• Optimization Framework 100%
• Source Machine 100%
• Deep Neural Network 40%
• Design-Space Exploration 40%
• Machine Learning Algorithm 20%
• Physical Design 20%
• Performance Simulation 20%
• Based Prediction Framework 20%
• Prediction Error 20%
• Application Specific Integrated Circuit 20%
• Deep Learning 20%