FP Quantization Solutions for Large Models
Summary of floating-point quantization (FPQ) for compressing large language models, covering FP formats, scaling, pre-shifted exponent bias, and 4-bit results.
Robotics & Artificial Intelligence
Patterns and Challenges in AI Chip Design
Analysis of AI in chip and system design: tradeoffs in energy, sparsity, memory and edge inference, plus architecture and EDA optimization.
Transformer Models for Machine Vision Object Detection
Analysis of transformer-based DETR for object detection, its CNN-transformer architecture, deployment on edge AI hardware, optimization, and security considerations.
Image Generation with Diffusion Models in PyTorch
Technical overview of diffusion models and practical PyTorch implementation for image generation, covering forward/reverse processes, training objectives, and sampling.
AI Workloads' Impact on Processor Design
Analysis of how AI reshapes processor design, exploring heterogeneous architectures, memory bandwidth, and PPA trade-offs for system-level optimization.
Top 10 Common Machine Learning Algorithms
Overview of ten common supervised learning algorithms—linear/logistic regression, SVM, trees, ensembles—and guidance on selection, assumptions, and dataset considerations.
Principles and Benefits of AI-Based Recommendation Systems
Technical overview of recommendation systems, models (collaborative, content, DLRM, Transformer) and GPU-accelerated frameworks like Merlin for large-scale training and inference
What Is Artificial Intelligence: Principles and Applications
Survey of artificial intelligence, machine learning and deep learning, technology, applications (healthcare, autonomous driving, materials) and future directions.
China's AI Development: Status and Challenges
Technical overview of artificial intelligence: history, deep learning advances, perception-to-cognition trends, and China's AI development strengths, gaps, and planning.
Fundamentals and Types of AI Foundation Software
China AI Foundation Software Market (2023) analysis of AI 2.0 and AI foundation software: value chain, vendor competitiveness, compute, data and business-model challenges.
Five Stages of AI Model Development
Overview of AI model development lifecycle covering model design, feature engineering, model training, validation, fusion and deployment for production ML systems.
How Is China Telecom Advancing Its Large Language Models?
China Telecom's approach to large models: organizational setup, data/compute resources, Xingchen model scaling, hallucination mitigation, multimodal and 3D digital-human results.
