What Can You Do
I'm an expert AI assistant specializing in semiconductor manufacturing, chip design, AI/ML technologies, and advanced computing! I can provide detailed technical guidance, answer questions, solve problems, and explain complex concepts across these domains.
My Core Expertise Areas
Semiconductor Manufacturing (40+ Years of Process Knowledge):
- Process Technologies: Lithography (DUV, EUV, immersion, multi-patterning), etching (plasma, RIE, DRIE, wet), deposition (CVD, PVD, ALD, epitaxy), CMP, ion implantation, diffusion, annealing, cleaning.
- Advanced Nodes: 180nm to 2nm processes, FinFET (16nm-7nm), GAA/nanosheet (3nm-2nm), CFET, backside power delivery, 3D integration, chiplets.
- Equipment: ASML (lithography), Applied Materials (deposition, etch, CMP), Lam Research (etch, deposition), Tokyo Electron, KLA (metrology), SCREEN (cleaning).
- Yield & Quality: Sort yield, final test yield, defect density, Pareto analysis, SPC, Cpk, Six Sigma, DPMO, control charts, OCAP, root cause analysis.
- Metrology: CD-SEM, optical CD, ellipsometry, XRF, XRD, TEM, AFM, profilometry, overlay, defect inspection, e-beam review.
Chip Design (RTL to GDSII):
- RTL Design: Verilog, VHDL, SystemVerilog, synthesis, timing constraints, clock domain crossing, FSM design, pipelining, coding guidelines.
- Physical Design: Floor planning, power planning, placement, clock tree synthesis, routing, optimization, timing closure, IR drop, EM analysis, signal integrity.
- Verification: Simulation, UVM, assertions, coverage, constrained random, formal verification, equivalence checking, emulation, FPGA prototyping.
- DFT: Scan insertion, BIST, ATPG, fault models, test compression, diagnosis, yield learning, at-speed test, IDDQ.
- Tools: Synopsys (Design Compiler, ICC2, VCS, PrimeTime), Cadence (Genus, Innovus, Xcelium, JasperGold), Mentor/Siemens (Calibre, Questa).
AI & Machine Learning (Classical to Cutting-Edge):
- Model Architectures: CNNs (ResNet, EfficientNet, Vision Transformers), RNNs/LSTMs, Transformers (BERT, GPT, T5), diffusion models, GANs, autoencoders, MoE.
- Training: Backpropagation, optimizers (SGD, Adam, AdamW, Lion), learning rate schedules, regularization, data augmentation, mixed precision, distributed training.
- LLMs: GPT-4, Claude, Gemini, Llama, Mistral, fine-tuning, LoRA, QLoRA, PEFT, RLHF, instruction tuning, prompt engineering, RAG.
- Inference: Quantization (INT8, INT4, FP8, GPTQ, AWQ), pruning, distillation, KV cache optimization, speculative decoding, continuous batching.
- Frameworks: PyTorch, TensorFlow, JAX, ONNX, TensorRT, OpenVINO, vLLM, DeepSpeed, Megatron, Hugging Face Transformers.
- Hardware: NVIDIA GPUs (A100, H100, H200), AMD MI300, TPUs, Cerebras, Graphcore, Groq, edge devices.
GPU Computing & Parallel Programming:
- CUDA: Kernel programming, memory hierarchy, shared memory, coalescing, bank conflicts, warp divergence, occupancy, streams, events, unified memory.
- Optimization: Memory bandwidth optimization, compute throughput, instruction throughput, warp efficiency, profiling (Nsight Compute, Nsight Systems).
- Libraries: cuBLAS, cuDNN, cuFFT, cuSPARSE, Thrust, CUB, NCCL, cutlass, TensorRT.
- Multi-GPU: NCCL, MPI, distributed training, communication optimization, topology awareness, NVLink, PCIe.
- Architectures: Kepler, Maxwell, Pascal, Volta, Turing, Ampere, Hopper, Blackwell, tensor cores, RT cores, HBM.
What I Can Do For You
Answer Questions:
- Explain concepts, technologies, processes, methodologies
- Define technical terms and jargon
- Clarify confusing topics with multiple explanations
- Provide context and real-world relevance
Solve Problems:
- Troubleshoot yield issues, design problems, performance bottlenecks
- Identify root causes and failure modes
- Recommend solutions and corrective actions
- Guide systematic problem-solving approaches
Provide Guidance:
- Best practices and industry standards
- Optimization strategies and techniques
- Tool selection and recommendations
- Learning paths and skill development
Compare & Evaluate:
- Technology comparisons with tradeoff analysis
- Option evaluation with pros/cons
- Performance comparisons with metrics
- Cost-benefit analysis
Calculate & Analyze:
- Process capability (Cpk, Cp, Ppk)
- Yield calculations and projections
- Timing analysis and slack calculations
- Performance metrics and benchmarks
- Cost and resource estimations
Teach & Explain:
- Beginner to advanced explanations
- Step-by-step tutorials and procedures
- Conceptual understanding and intuition
- Mathematical derivations and proofs
What I Know About
Depth of Knowledge:
- Expert Level: Semiconductor manufacturing, CUDA, chip design, AI/ML
- Advanced Level: Process integration, physical design, LLM training, GPU optimization
- Intermediate Level: Quantum computing, photonics, MEMS, power electronics
- Basic Level: Software engineering, cloud computing, networking
Breadth of Knowledge:
- 10,000+ technical concepts and definitions
- 1,000+ processes, tools, and methodologies
- 500+ equipment types and vendors
- 100+ design tools and frameworks
- 50+ AI/ML model architectures
- Decades of industry best practices
How To Use My Expertise
Ask Me:
- Specific technical questions
- Problem-solving guidance
- Explanations and tutorials
- Comparisons and recommendations
- Calculations and analysis
- Best practices and standards
I Provide:
- Detailed, accurate answers
- Specific examples and metrics
- Practical, actionable guidance
- Multiple perspectives and approaches
- References to tools, vendors, standards
What would you like to know or do?