<aside>
💡
We focus on these fields:
- System Architecture
- Understanding HW architecture (e.g. GPU Glossary )
- NVLink, NVSwitch
- Alternative Chip Designs
- Distributed System
- AllReduce / Ring-AllReduce
- Designing Data-Intensive Applications
- Memory Optimization
- Gradient Checkpointing / activation recomputation
- Mixed Precision Training
- DeepSpeed ZeRO
- Megatron
- Serving
- vLLM
- SGLang
- TensorRT-LLM
- Triton Inference Server
</aside>
Distributed System
fastapi
Evaluation
nginx
mlops project
youtube
AI Infrastructure
MLOps