https://lilinji.mintlify.app/01_性能工程与系统前置/01_AI_Infra工程师的C++与Python底层必修课 2026-09-28T13:43:25.172Z https://lilinji.mintlify.app/01_性能工程与系统前置/02_AI_Infra数学基础 2026-09-28T13:43:25.167Z https://lilinji.mintlify.app/01_性能工程与系统前置/03_计算机体系结构前置_CPU存储层次_CacheLine与NUMA架构 2026-09-28T13:43:25.161Z https://lilinji.mintlify.app/01_性能工程与系统前置/04_Linux系统基线与性能分析核心工具箱 2026-09-28T13:43:25.167Z https://lilinji.mintlify.app/01_性能工程与系统前置/05_PyTorch底层系统机制_Tensor内存布局_Stride与Contiguous 2026-09-28T13:43:25.165Z https://lilinji.mintlify.app/01_性能工程与系统前置/06_PyTorch计算图与Autograd显存生命周期 2026-09-28T13:43:25.174Z https://lilinji.mintlify.app/01_性能工程与系统前置/07_PyTorch异步执行流_CUDA_Stream与CUDA_Graph原理 2026-09-28T13:43:25.170Z https://lilinji.mintlify.app/01_性能工程与系统前置/08_Transformer系统全景与算子执行图深度拆解 2026-09-28T13:43:25.163Z https://lilinji.mintlify.app/01_性能工程与系统前置/09_网络通信基础与集合通信Collective原语直觉 2026-09-28T13:43:25.163Z https://lilinji.mintlify.app/01_性能工程与系统前置/10_AI_Infra性能工程方法论与三账本_FLOPs_显存_通信 2026-09-28T13:43:25.162Z https://lilinji.mintlify.app/01_性能工程与系统前置/11_PyTorch_Profiler实战_手把手带你做一次性能体检 2026-09-28T13:43:25.164Z https://lilinji.mintlify.app/01_性能工程与系统前置/AI Infra 大话西游实战之Transformer小白学习笔记（一） 2026-09-28T13:43:25.168Z https://lilinji.mintlify.app/01_性能工程与系统前置/README 2026-09-28T13:43:02.121Z https://lilinji.mintlify.app/02_GPU硬件与集群互联/12_GPU执行与存储体系_SM_Warp_TensorCore_HBM_Roofline 2026-09-28T13:43:02.121Z https://lilinji.mintlify.app/02_GPU硬件与集群互联/13_通信第一性原理与性能模型_AlphaBeta_Latency_Bandwidth 2026-09-28T13:43:02.116Z https://lilinji.mintlify.app/02_GPU硬件与集群互联/14_GPU节点与集群硬件拓扑_PCIe_NVLink_NVSwitch_Clos_Rail 2026-09-28T13:43:02.112Z https://lilinji.mintlify.app/02_GPU硬件与集群互联/15_RDMA与GPUDirect_RDMA_QP_WQE_CQ_MR_ZeroCopy 2026-09-28T13:43:02.130Z https://lilinji.mintlify.app/02_GPU硬件与集群互联/16_机内数据搬运_LSU_TMA_CopyEngine_mbarrier_IPC 2026-09-28T13:43:02.131Z https://lilinji.mintlify.app/02_GPU硬件与集群互联/17_机间数据搬运_GPU_Initiated_RDMA_IBRC_OneHop 2026-09-28T13:43:02.131Z https://lilinji.mintlify.app/02_GPU硬件与集群互联/18_NCCL通信库与Collective算子_Ring_Tree_NVLS_NVSHMEM 2026-09-28T13:43:02.107Z https://lilinji.mintlify.app/02_GPU硬件与集群互联/19_Compute_Communication_Overlap_Stream_Bucket_Chunk_WarpSpec 2026-09-28T13:43:02.130Z https://lilinji.mintlify.app/02_GPU硬件与集群互联/20_Benchmark_监控与AI集群故障诊断_Xid_SlowNode_NCCL_Hang 2026-09-28T13:43:02.113Z https://lilinji.mintlify.app/02_GPU硬件与集群互联/21_GPU硬件全景与多元异构算力_NVIDIA代际演进_国产芯片与生态选型 2026-09-28T13:43:02.103Z https://lilinji.mintlify.app/02_GPU硬件与集群互联/README 2026-09-28T13:43:02.129Z https://lilinji.mintlify.app/03_CUDA与算子性能优化/22_CUDA编程模型与内存层次_Thread_Block_Coalescing 2026-09-28T13:43:02.106Z https://lilinji.mintlify.app/03_CUDA与算子性能优化/23_经典算子优化_Reduce与Softmax_OnlineSoftmax_Fused 2026-09-28T13:43:02.122Z https://lilinji.mintlify.app/03_CUDA与算子性能优化/24_矩阵乘法核心_GEMM优化与TensorCore思维 2026-09-28T13:43:02.110Z https://lilinji.mintlify.app/03_CUDA与算子性能优化/25_注意力算子深度优化_TiledAttention_FlashAttention_Triton 2026-09-28T13:43:02.118Z https://lilinji.mintlify.app/03_CUDA与算子性能优化/README 2026-09-28T13:43:02.109Z https://lilinji.mintlify.app/04_大模型架构与显存建模/26_Decoder_only架构深度拆解_MHA_MQA_GQA_RoPE_SwiGLU 2026-09-28T13:43:02.129Z https://lilinji.mintlify.app/04_大模型架构与显存建模/27_训练与推理显存账本_FLOPs推导与KV_Cache容量规划 2026-09-28T13:43:02.117Z https://lilinji.mintlify.app/04_大模型架构与显存建模/README 2026-09-28T13:43:02.119Z https://lilinji.mintlify.app/05_大模型分布式训练/28_DDP分布式数据并行与通信计算重叠 2026-09-28T13:43:02.117Z https://lilinji.mintlify.app/05_大模型分布式训练/29_显存分片技术_ZeRO_1_2_3与PyTorch_FSDP深度剖析 2026-09-28T13:43:02.113Z https://lilinji.mintlify.app/05_大模型分布式训练/30_模型并行技术_TP_PP_SP_CP与硬件拓扑映射 2026-09-28T13:43:02.122Z https://lilinji.mintlify.app/05_大模型分布式训练/31_3D混合并行_MoE专家并行与大规模训练故障排查 2026-09-28T13:43:02.123Z https://lilinji.mintlify.app/05_大模型分布式训练/README 2026-09-28T13:43:02.064Z https://lilinji.mintlify.app/06_LLM推理系统与性能工程/32_LLM推理两阶段_Prefill_Decode与核心SLO指标 2026-09-28T13:43:02.125Z https://lilinji.mintlify.app/06_LLM推理系统与性能工程/33_KV_Cache管理演进_PagedAttention与ContinuousBatching 2026-09-28T13:43:02.066Z https://lilinji.mintlify.app/06_LLM推理系统与性能工程/34_vLLM系统架构剖析与高阶加速技术 2026-09-28T13:43:02.105Z https://lilinji.mintlify.app/06_LLM推理系统与性能工程/35_分布式推理与PD分离架构_Prefill_Decode_Disaggregation 2026-09-28T13:43:02.106Z https://lilinji.mintlify.app/06_LLM推理系统与性能工程/36_在线推理容量规划_SLO治理与生产故障诊断 2026-09-28T13:43:02.128Z https://lilinji.mintlify.app/06_LLM推理系统与性能工程/README 2026-09-28T13:43:02.115Z https://lilinji.mintlify.app/07_云原生AI平台与生产工程/37_GPU容器化与Kubernetes_Device_Plugin_DRA 2026-09-28T13:43:02.114Z https://lilinji.mintlify.app/07_云原生AI平台与生产工程/38_AI任务调度与多租户隔离_Gang_Kueue_HAMi_GPUSharing 2026-09-28T13:43:02.064Z https://lilinji.mintlify.app/07_云原生AI平台与生产工程/39_生产级AI_Serving高可用_可观测与SLO治理 2026-09-28T13:43:02.111Z https://lilinji.mintlify.app/07_云原生AI平台与生产工程/40_大模型分布式存储_Checkpoint与缓存体系架构 2026-09-28T13:43:02.108Z https://lilinji.mintlify.app/07_云原生AI平台与生产工程/README 2026-09-28T13:43:02.063Z https://lilinji.mintlify.app/08_Post-Training与相邻Infra/41_SFT_LoRA与RLHF系统负载与资源评估 2026-09-28T13:43:02.132Z https://lilinji.mintlify.app/08_Post-Training与相邻Infra/42_RAG与Agent基础设施_VectorDB_Context_Sandbox 2026-09-28T13:43:02.066Z https://lilinji.mintlify.app/08_Post-Training与相邻Infra/43_NPU_DPU与异构加速器硬件抽象 2026-09-28T13:43:02.110Z https://lilinji.mintlify.app/08_Post-Training与相邻Infra/44_AI_Infra大厂全真面试演练与系统设计题库 2026-09-28T13:43:02.123Z https://lilinji.mintlify.app/08_Post-Training与相邻Infra/README 2026-09-28T13:43:02.068Z https://lilinji.mintlify.app/09_Capstone综合实战项目/01_Capstone_分布式训练Scaling_Study实战 2026-09-28T13:43:02.125Z https://lilinji.mintlify.app/09_Capstone综合实战项目/02_Capstone_生产级LLM_Serving吞吐与延迟优化实战 2026-09-28T13:43:02.124Z https://lilinji.mintlify.app/09_Capstone综合实战项目/03_Capstone_CUDA_Attention算子极致性能优化实战 2026-09-28T13:43:02.065Z https://lilinji.mintlify.app/09_Capstone综合实战项目/04_Capstone_企业级AI_Platform统一调度与集群架构RFC 2026-09-28T13:43:02.115Z https://lilinji.mintlify.app/09_Capstone综合实战项目/05_Capstone_FlashAttention全景与Token生成端到端全链路优化实战 2026-09-28T13:43:02.108Z https://lilinji.mintlify.app/09_Capstone综合实战项目/README 2026-09-28T13:43:02.065Z https://lilinji.mintlify.app/content-types 2026-09-28T13:43:02.067Z https://lilinji.mintlify.app 2026-09-28T13:43:02.067Z https://lilinji.mintlify.app/style-guide 2026-09-28T13:43:02.124Z