Inference 资料

1. P010
   PromptTuner: SLO-Aware Elastic System for LLM Prompt Tuning

   arXiv:2603.05087 · Scholar 收录
2. P013
   DAWN: Dependency-Aware Fast Inference for Diffusion LLMs

   arXiv:2602.06953 · Scholar 收录
3. P018
   Inference-time Alignment via Sparse Junction Steering

   arXiv:2602.21215 · Scholar 收录
4. P022
   Semantic-Aware Scheduling for GPU Clusters with Large Language Models

   arXiv:2510.03334 · Scholar 收录
5. P031
   DDiT: Dynamic Resource Allocation for Diffusion Transformer Model Serving

   arXiv:2506.13497 · Scholar 收录
6. P047
   UniTG: A Unified System for Efficient and Seamless Textual Graph Learning

   International Conference on Very Large Data Bases (VLDB), August, 2026
7. P054
   SpecForge: A Flexible and Efficient Open-Source Training Framework for Speculative Decoding

   International Conference on Machine Learning (ICML), July, 2026
8. P055
   CONCUR: High-Throughput Agentic Batch Inference of LLM via Congestion-Based Concurrency Control

   International Conference on Machine Learning (ICML), July, 2026
9. P056
   DSB: Dynamic Sliding Block Scheduling for Diffusion LLMs

   International Conference on Machine Learning (ICML), July, 2026
10. P064
    SPPO: Making Million-Token LLM Training Practical on Modest GPU Clusters

    ACM International Conference on Supercomputing (ICS), July, 2026
11. P070
    Di-PS: System-Algorithm Co-Design for Asynchronous and Heterogeneous Cross-cluster LLM Training at Scale

    USENIX Symposium on Networked Systems Design and Implementation (NSDI), May, 2026
12. P072
    ReSpec: Towards Optimizing Speculative Decoding in Reinforcement Learning Systems

    Annual Conference on Machine Learning and Systems (MLSys), May, 2026
13. P077
    DSA: Efficient Inference For Video Generation Models via Distributed Sparse Attention

    International Conference on Learning Representations (ICLR), April, 2026
14. P097
    Impact-driven Context Filtering For Cross-file Code Completion

    Conference on Language Modeling (COLM), October, 2025
15. P114
    Rethinking Key-Value Cache Compression Techniques for Large Language Model Serving

    Annual Conference on Machine Learning and Systems (MLSys), May, 2025
16. P129
    TorchGT: A Holistic System for Large-scale Graph Transformer Training

    International Conference for High Performance Computing, Networking, Storage, and Analysis (SC), November, 2024
17. P151
    Lins: Reducing Communication Overhead of ZeRO for Efficient LLM Training

    IEEE/ACM International Symposium on Quality of Service (IWQoS), June, 2024
18. P152
    Ymir: A Scheduler for Foundation Model Fine-tuning Workloads in Datacenters

    ACM International Conference on Supercomputing (ICS), June, 2024
19. P153
    AutoSched: An Adaptive Self-configured Framework for Scheduling Deep Learning Training Workloads

    ACM International Conference on Supercomputing (ICS), June, 2024
20. P158
    Sylvie: 3D-adaptive and Universal System for Large-scale Graph Neural Network Training

    IEEE International Conference on Data Engineering (ICDE), May, 2024
21. P164
    Characterization of Large Language Model Development in the Datacenter

    USENIX Symposium on Networked Systems Design and Implementation (NSDI), April, 2024
22. P177
    Hydro: Surrogate-based Hyperparameter Tuning Service in Datacenters

    USENIX Symposium on Operating Systems Design and Implementation (OSDI), July, 2023
23. P187
    Lucid: A Non-Intrusive, Scalable and Interpretable Scheduler for Deep Learning Training Jobs

    ACM International Conference on Architectural Support for Programming Languages and Operating Systems (ASPLOS), Distinguished Paper Award , March, 2023
24. P193
    Titan: A Scheduler for Foundation Model Fine-tuning Workloads

    ACM Symposium on Cloud Computing (SoCC), November, 2022
25. P194
    Tear Up the Bubble Boom: Lessons Learned From a Deep Learning Research and Development Cluster

    IEEE International Conference on Computer Design (ICCD), October, 2022
26. P208
    CHRONUS: A Novel Deadline-aware Scheduler for Deep Learning Training Jobs

    ACM Symposium on Cloud Computing (SoCC), November, 2021
27. P209
    Characterization and Prediction of Deep Learning Workloads in Large-Scale GPU Datacenters

    International Conference for High Performance Computing, Networking, Storage, and Analysis (SC), November, 2021
28. P239
    Collaborative Inference and Learning between Edge SLMs and Cloud LLMs: A Survey of Algorithms, Execution, and Open Challenges

    Accepted by ACM Computing Surveys
29. P264
    ICEFROG: A Layer-Elastic Scheduling System for Deep Learning Training in GPU Clusters

    IEEE Transactions on Parallel and Distributed Systems, Volume: 36, Issue: 6, June 2025
30. P275
    UniSched: A Unified Scheduler for Deep Learning Training Jobs with Different User Demands

    IEEE Transactions on Computers, Volume: 73, Issue: 6, June 2024
31. P282
    Deep Learning Workload Scheduling in GPU Datacenters: A Survey

    ACM Computing Surveys, Volume: 56, Issue: 6, January 2024
32. P300
    ASTRAEA: A Fair Deep Learning Scheduler for Multi-tenant GPU Clusters

    IEEE Transactions on Parallel and Distributed Systems, Volume: 33, Issue: 11, November 2022
33. O01
    Omni serving 到底是什么

    Omni Inference · 基础与工作负载 · 核验 2026-08-07
34. O02
    模型图、模态状态与输出 codec

    Omni Inference · 基础与工作负载 · 核验 2026-08-07
35. O03
    2023–2026 Omni serving 论文谱系

    Omni Inference · 论文与系统谱系 · 核验 2026-08-08
36. O04
    vLLM-Omni 架构与源码请求路径

    Omni Inference · vLLM-Omni · 核验 2026-08-07
37. O05
    vLLM-Omni 调度、缓存、Diffusion 与 Connector

    Omni Inference · vLLM-Omni · 核验 2026-08-07
38. O06
    vLLM-Omni 论文、部署与实验边界

    Omni Inference · vLLM-Omni · 核验 2026-08-07
39. O07
    SGLang-Omni 架构与源码请求路径

    Omni Inference · SGLang-Omni · 核验 2026-08-07
40. O08
    SGLang-Omni 调度、缓存、通信与部署

    Omni Inference · SGLang-Omni · 核验 2026-08-07
41. O09
    vLLM-Omni 与 SGLang-Omni：逐层比较

    Omni Inference · 横向分析 · 核验 2026-08-07
42. O10
    状态、背压、多租户与故障语义

    Omni Inference · 横向分析 · 核验 2026-08-07
43. O11
    可复现评测与容量实验

    Omni Inference · 研究方法 · 核验 2026-08-07
44. O12
    系统研究议题与实验路线

    Omni Inference · 研究方法 · 核验 2026-08-07
45. 001
    Distributed Deep Neural Networks Over the Cloud, the Edge and End Devices (ICDCS 2017)

    Cloud-Edge Collaboration · Automatic Graph Partitioning · 2017
46. 002
    Dynamic Adaptive DNN Surgery for Inference Acceleration on the Edge (INFOCOM 2019)

    Cloud-Edge Collaboration · Automatic Graph Partitioning · 2019
47. 003
    Neurosurgeon: Collaborative Intelligence Between the Cloud and Mobile Edge

    Cloud-Edge Collaboration · Automatic Graph Partitioning · 2017
48. 004
    SPINN: Synergistic Progressive Inference of Neural Networks over Device and Cloud (MobiCom 2020)

    Cloud-Edge Collaboration · Automatic Graph Partitioning · 2020
49. 005
    Walle: An End-to-End, General-Purpose, and Large-Scale Production System for Device-Cloud Collaborative Machine Learning

    Cloud-Edge Collaboration · Framework · 2022
50. 006
    A case for disaggregation of ML data processing (arXiv 2210.14826)

    Data Processing · Uncategorized · 2022-10
51. 007
    Disaggregating ML Input Data Processing at Scale

    Data Processing · Uncategorized · 2023
52. 008
    GoldMiner: Elastic Scaling of Training Data Pre-Processing Pipelines for Deep Learning (SIGMOD 2023)

    Data Processing · Uncategorized · 2023
53. 009
    Pecan: Cost-Efficient ML Data Preprocessing with Automatic Transformation Ordering and Hybrid Placement

    Data Processing · Uncategorized · 2024
54. 010
    Understanding Data Storage and Ingestion for Large-Scale Deep Recommendation Model Training (ISCA 2022)

    Data Processing · Uncategorized · 2022
55. 011
    MLIR: Scaling Compiler Infrastructure for Domain Specific Computation (CGO 2021)

    Deep Learning Compiler · System Architecture · 2021
56. 012
    TVM: An Automated End-to-End Optimizing Compiler for Deep Learning

    Deep Learning Compiler · System Architecture · 2018
57. 013
    AStitch: Enabling a New Multi-dimensional Optimization Space for Memory-Intensive ML Training and Inference on Modern SIMT Architectures (ASPLOS 2022)

    Deep Learning Compiler · Tensor Program Generation / General Tensor Program Generation · 2022
58. 014
    Ansor: Generating High-Performance Tensor Programs for Deep Learning

    Deep Learning Compiler · Tensor Program Generation / General Tensor Program Generation · 2020
59. 015
    Cocktailer: Analyzing and Optimizing Dynamic Control Flow in Deep Learning

    Deep Learning Compiler · Tensor Program Generation / General Tensor Program Generation · 2023
60. 016
    EINNET: Optimizing Tensor Programs with Derivation-Based Transformations

    Deep Learning Compiler · Tensor Program Generation / General Tensor Program Generation · 2023
61. 017
    Effectively Scheduling Computational Graphs of Deep Neural Networks toward Their Domain-Specific Accelerators

    Deep Learning Compiler · Tensor Program Generation / General Tensor Program Generation · 2023
62. 018
    Welder: Scheduling Deep Learning Memory Access via Tile-graph

    Deep Learning Compiler · Tensor Program Generation / General Tensor Program Generation · 2023
63. 019
    Event Tensor: A Unified Abstraction for Compiling Dynamic Megakernel

    Deep Learning Compiler · Tensor Program Generation / Megakernel Compilation · 2026
64. 020
    Mirage Persistent Kernel: A Compiler and Runtime for Mega-Kernelizing Tensor Programs (arXiv:2512.22219)

    Deep Learning Compiler · Tensor Program Generation / Megakernel Compilation · 2025-12
65. 021
    An Introduction to Computational Networks and the Computational Network Toolkit (MSR-TR-2014-112)

    Deep Learning Framework · Uncategorized · 2014-01
66. 022
    Caffe: Convolutional Architecture for Fast Feature Embedding (arXiv 1408.5093)

    Deep Learning Framework · Uncategorized · 2014-08
67. 023
    Jittor: a novel deep learning framework with meta-operators and unified graph execution (Science China Information Sciences 2020)

    Deep Learning Framework · Uncategorized · 2020-02
68. 024
    MXNet: A Flexible and Efficient Machine Learning Library for Heterogeneous Distributed Systems (NIPS 2016 Workshop on MLSys)

    Deep Learning Framework · Uncategorized · 2016
69. 025
    OneFlow: Redesign the Distributed Deep Learning Framework from Scratch (arXiv 2110.15032)

    Deep Learning Framework · Uncategorized · 2021-10
70. 026
    Pathways: Asynchronous Distributed Dataflow for ML (MLSys 2022)

    Deep Learning Framework · Uncategorized · 2022
71. 027
    PyTorch: An Imperative Style, High-Performance Deep Learning Library (NeurIPS 2019)

    Deep Learning Framework · Uncategorized · 2019
72. 028
    TensorFlow: A System for Large-Scale Machine Learning (OSDI 2016)

    Deep Learning Framework · Uncategorized · 2016
73. 029
    XDL: An Industrial Deep Learning Framework for High-dimensional Sparse Data (DLP-KDD 2019)

    Deep Learning Framework · Uncategorized · 2019
74. 030
    EasyScale: Elastic Training with Consistent Accuracy and Improved Utilization on GPUs

    Deep Learning Training · Elastic Training · 2023
75. 031
    A Unified Architecture for Accelerating Distributed DNN Training in Heterogeneous GPU/CPU Clusters

    Deep Learning Training · Optimizing Network Communication · 2020
76. 032
    Hanayo: Harnessing Wave-like Pipeline Parallelism for Enhanced Large Model Training Efficiency

    Deep Learning Training · Parallelism · 2023
77. 033
    One weird trick for parallelizing convolutional neural networks (arXiv 1404.599)

    Deep Learning Training · Parallelism · 2014-04
78. 034
    Supporting Very Large Models using Automatic Dataflow Graph Partitioning

    Deep Learning Training · Parallelism · 2019
79. 035
    Echo: Compiler-based GPU Memory Footprint Reduction for LSTM RNN Training (ISCA 2020)

    Deep Learning Training · Reduce GPU Memory Footprints / Compression · 2020
80. 036
    Gist: Efficient Data Encoding for Deep Neural Network Training (ISCA 2018)

    Deep Learning Training · Reduce GPU Memory Footprints / Compression · 2018-04
81. 037
    Gandiva: Introspective Cluster Scheduling for Deep Learning

    Deep Learning Training · Reduce GPU Memory Footprints / GPU Sharing · 2018
82. 038
    Salus: Fine-Grained GPU Sharing Primitives for Deep Learning Applications

    Deep Learning Training · Reduce GPU Memory Footprints / GPU Sharing · 2020
83. 039
    Zico: Efficient GPU Memory Sharing for Concurrent DNN Training

    Deep Learning Training · Reduce GPU Memory Footprints / GPU Sharing · 2021
84. 040
    Capuchin: Tensor-based GPU Memory Management for Deep Learning

    Deep Learning Training · Reduce GPU Memory Footprints / Tensor Swapping / Recomputation · 2020
85. 041
    Checkmate: Breaking the Memory Wall with Optimal Tensor Rematerialization

    Deep Learning Training · Reduce GPU Memory Footprints / Tensor Swapping / Recomputation · 2020
86. 042
    SuperNeurons: Dynamic GPU Memory Management for Training Deep Neural Networks (PPoPP 2018)

    Deep Learning Training · Reduce GPU Memory Footprints / Tensor Swapping / Recomputation · 2018
87. 043
    SwapAdvisor: Pushing Deep Learning Beyond the GPU Memory Limit via Smart Swapping

    Deep Learning Training · Reduce GPU Memory Footprints / Tensor Swapping / Recomputation · 2020
88. 044
    Training Deep Nets with Sublinear Memory Cost (arXiv 1604.06174)

    Deep Learning Training · Reduce GPU Memory Footprints / Tensor Swapping / Recomputation · 2016
89. 045
    vDNN: Virtualized Deep Neural Networks for Scalable, Memory-Efficient Neural Network Design (MICRO 2016)

    Deep Learning Training · Reduce GPU Memory Footprints / Tensor Swapping / Recomputation · 2016
90. 046
    CAT-DM: Controllable Accelerated Virtual Try-on with Diffusion Model

    Diffusion Models · Diffusion Model Serving / Image Editing · 2024
91. 047
    FlashPS: Efficient Generative Image Editing with Mask-aware Caching and Scheduling

    Diffusion Models · Diffusion Model Serving / Image Editing · 2026
92. 048
    Approximate Caching for Efficiently Serving Text-to-Image Diffusion Models

    Diffusion Models · Diffusion Model Serving / Image Generation · 2024
93. 049
    Cache Me if You Can: Accelerating Diffusion Models through Block Caching

    Diffusion Models · Diffusion Model Serving / Image Generation · 2024
94. 050
    DeepCache: Accelerating Diffusion Models for Free

    Diffusion Models · Diffusion Model Serving / Image Generation · 2024
95. 051
    DiFlow: A System for Micro-Serving Text-to-Image Diffusion Workflows

    Diffusion Models · Diffusion Model Serving / Image Generation · 2026
96. 052
    DistriFusion: Distributed Parallel Inference for High-Resolution Diffusion Models

    Diffusion Models · Diffusion Model Serving / Image Generation · 2024
97. 053
    Katz: Efficient Workflow Serving for Diffusion Models with Many Adapters

    Diffusion Models · Diffusion Model Serving / Image Generation · 2025
98. 054
    MixFusion: A Patch-Level Parallel Serving System for Mixed-Resolution Diffusion Models

    Diffusion Models · Diffusion Model Serving / Image Generation · 2026
99. 055
    PipeFusion: Displaced Patch Pipeline Parallelism for Inference of Diffusion Transformer Models (arXiv:2405.14430)

    Diffusion Models · Diffusion Model Serving / Image Generation · 2024-05
100. 056
     xDiT: an Inference Engine for Diffusion Transformers (DiTs) with Massive Parallelism (arXiv:2411.01738)

     Diffusion Models · Diffusion Model Serving / Image Generation · 2024-11
101. 057
     Fast Video Generation with Sliding Tile Attention

     Diffusion Models · Diffusion Model Serving / Video Generation · 2025
102. 058
     FlexCache: Flexible Approximate Cache System for Video Diffusion (arXiv:2501.04012)

     Diffusion Models · Diffusion Model Serving / Video Generation · 2025-01
103. 059
     DiffusionPipe: Training Large Diffusion Models with Efficient Pipelines

     Diffusion Models · Diffusion Model Training · 2024
104. 060
     Cambricon-D: Full-Network Differential Acceleration for Diffusion Models

     Diffusion Models · Domain-Specific Accelerator (DSA) · 2024
105. 061
     X-Adapter: Adding Universal Compatibility of Plugins for Upgraded Diffusion Model

     Diffusion Models · Supporting Add-on Modules · 2024
106. 062
     DisaggRec: Architecting Disaggregated Systems for Large-Scale Personalized Recommendation (arXiv 2212.00939)

     Deep Learning Recommendation Model (DLRM) · DLRM Inference · 2022
107. 063
     Accelerating Neural Recommendation Training with Embedding Scheduling

     Deep Learning Recommendation Model (DLRM) · DLRM Training · 2024
108. 064
     Bagpipe: Accelerating Deep Recommendation Model Training

     Deep Learning Recommendation Model (DLRM) · DLRM Training · 2023
109. 065
     Heterogeneous Acceleration Pipeline for Recommendation System Training

     Deep Learning Recommendation Model (DLRM) · DLRM Training · 2024
110. 066
     EVStore: Storage and Caching Capabilities for Scaling Embedding Tables in Deep Recommendation Systems

     Deep Learning Recommendation Model (DLRM) · GPU Cache · 2023
111. 067
     UGache: A Unified GPU Cache for Embedding-based Deep Learning

     Deep Learning Recommendation Model (DLRM) · GPU Cache · 2023
112. 068
     Ekko: A Large-Scale Deep Learning Recommender System with Low-Latency Model Update

     Deep Learning Recommendation Model (DLRM) · Model Update · 2022
113. 069
     AdaEmbed: Adaptive Embedding for Large-Scale Recommendation Models

     Deep Learning Recommendation Model (DLRM) · Pruning · 2023
114. 070
     CherryPick: Adaptively Unearthing the Best Cloud Configurations for Big Data Analytics

     Hyper-Parameter Tuning (HPO) · HPO for Systems · 2017
115. 071
     Google Vizier: A Service for Black-Box Optimization (KDD 2017)

     Hyper-Parameter Tuning (HPO) · HPO for Systems · 2017
116. 072
     Morphling: Fast, Near-Optimal Auto-Configuration for Cloud-Native Model Serving

     Hyper-Parameter Tuning (HPO) · HPO for Systems · 2021
117. 073
     Selecting the Best VM across Multiple Public Clouds: A Data-Driven Performance Modeling Approach

     Hyper-Parameter Tuning (HPO) · HPO for Systems · 2017
118. 074
     Elastic Hyperparameter Tuning on the Cloud

     Hyper-Parameter Tuning (HPO) · Optimizing HPO Workloads · 2021
119. 075
     Hydro: Surrogate-Based Hyperparameter Tuning Service in Datacenters

     Hyper-Parameter Tuning (HPO) · Optimizing HPO Workloads · 2023
120. 076
     RubberBand: Cloud-based Hyperparameter Tuning

     Hyper-Parameter Tuning (HPO) · Optimizing HPO Workloads · 2021
121. 077
     PUZZLE: Efficiently Aligning Large Language Models through Light-Weight Context Switch

     Large Language Model (LLM) · LLM Alignment · 2024
122. 078
     ThunderAgent: A Fast, Simple, and Program-Aware Agentic Inference System

     Large Language Model (LLM) · LLM Inference / Agentic Inference · 2026
123. 079
     LMPrefill: An Inference Engine for Prefill-only Workloads in Large Language Model Applications

     Large Language Model (LLM) · LLM Inference / Chunked Prefill · 2025
124. 080
     Taming Throughput-Latency Tradeoff in LLM Inference with Sarathi-Serve

     Large Language Model (LLM) · LLM Inference / Chunked Prefill · 2024
125. 081
     Tender: Accelerating Large Language Models via Tensor Decomposition and Runtime Requantization

     Large Language Model (LLM) · LLM Inference / Compression · 2024
126. 082
     Fairness in Serving Large Language Models

     Large Language Model (LLM) · LLM Inference / Fairness · 2024
127. 083
     Locality-aware Fair Scheduling in LLM Serving (arXiv:2501.14312)

     Large Language Model (LLM) · LLM Inference / Fairness · 2025-01
128. 084
     Cauchy: A Cost-Efficient LLM Serving System through Adaptive Heterogeneous Deployment (SoCC 2025)

     Large Language Model (LLM) · LLM Inference / Heterogeneous Deployment · 2025
129. 085
     Coral: Cost-Efficient Multi-LLM Serving over Heterogeneous Cloud GPUs (arXiv:2605.04357)

     Large Language Model (LLM) · LLM Inference / Heterogeneous Deployment · 2026-05
130. 086
     Demystifying Cost-Efficiency in LLM Serving over Heterogeneous GPUs (arXiv:2502.00722)

     Large Language Model (LLM) · LLM Inference / Heterogeneous Deployment · 2025-02
131. 087
     HexGen-2: Disaggregated Generative Inference of LLMs in Heterogeneous Environment (ICLR 2025)

     Large Language Model (LLM) · LLM Inference / Heterogeneous Deployment · 2025
132. 088
     HexGen: Generative Inference of Foundation Model over Heterogeneous Decentralized Environment

     Large Language Model (LLM) · LLM Inference / Heterogeneous Deployment · 2024
133. 089
     SageServe: Optimizing LLM Serving on Cloud Data Centers with Forecast Aware Auto-Scaling (SIGMETRICS Abstracts 2026)

     Large Language Model (LLM) · LLM Inference / Heterogeneous Deployment · 2026
134. 090
     SpotServe: Serving Generative Large Language Models on Preemptible Instances

     Large Language Model (LLM) · LLM Inference / Heterogeneous Deployment · 2024
135. 091
     ALISA: Accelerating Large Language Model Inference via Sparsity-Aware KV Caching

     Large Language Model (LLM) · LLM Inference / KV Cache Management · 2024
136. 092
     CacheBlend: Fast Large Language Model Serving for RAG with Cached Knowledge Fusion

     Large Language Model (LLM) · LLM Inference / KV Cache Management · 2025
137. 093
     CacheGen: KV Cache Compression and Streaming for Fast Large Language Model Serving

     Large Language Model (LLM) · LLM Inference / KV Cache Management · 2024
138. 094
     DroidSpeak: KV Cache Sharing Across Fine-tuned Model Variants

     Large Language Model (LLM) · LLM Inference / KV Cache Management · 2026
139. 095
     Efficient Memory Management for Large Language Model Serving with PagedAttention

     Large Language Model (LLM) · LLM Inference / KV Cache Management · 2023
140. 096
     Jenga: Effective Memory Management for Serving LLM with Heterogeneity

     Large Language Model (LLM) · LLM Inference / KV Cache Management · 2025
141. 097
     Prompt Cache: Modular Attention Reuse for Low-Latency Inference

     Large Language Model (LLM) · LLM Inference / KV Cache Management · 2024
142. 098
     Parrot: Efficient Serving of LLM-based Applications with Semantic Variable

     Large Language Model (LLM) · LLM Inference / LLM-based Applications · 2024
143. 099
     SGLang: Efficient Execution of Structured Language Model Programs

     Large Language Model (LLM) · LLM Inference / LLM-based Applications · 2024
144. 100
     Teola: Towards End-to-End Optimization of LLM-based Applications

     Large Language Model (LLM) · LLM Inference / LLM-based Applications · 2025
145. 101
     CaraServe: CPU-Assisted and Rank-Aware LoRA Serving for Generative LLM Inference (arXiv:2401.11240)

     Large Language Model (LLM) · LLM Inference / LoRA Serving · 2024-01
146. 102
     Punica: Multi-Tenant LoRA Serving

     Large Language Model (LLM) · LLM Inference / LoRA Serving · 2024
147. 103
     S-LoRA: Serving Thousands of Concurrent LoRA Adapters

     Large Language Model (LLM) · LLM Inference / LoRA Serving · 2024
148. 104
     dLoRA: Dynamically Orchestrating Requests and Adapters for LoRA LLM Serving

     Large Language Model (LLM) · LLM Inference / LoRA Serving · 2024
149. 105
     FlexGen: High-Throughput Generative Inference of Large Language Models with a Single GPU

     Large Language Model (LLM) · LLM Inference / Offloading · 2023
150. 106
     LLM in a flash: Efficient Large Language Model Inference with Limited Memory (arXiv 2312.11514)

     Large Language Model (LLM) · LLM Inference / Offloading · 2023-12
151. 107
     AlpaServe: Statistical Multiplexing with Model Parallelism for Deep Learning Serving

     Large Language Model (LLM) · LLM Inference / Parallelism and Partitioning · 2023
152. 108
     DeepSpeed-Inference: Enabling Efficient Inference of Transformer Models at Unprecedented Scale

     Large Language Model (LLM) · LLM Inference / Parallelism and Partitioning · 2022
153. 109
     Efficiently Scaling Transformer Inference

     Large Language Model (LLM) · LLM Inference / Parallelism and Partitioning · 2023
154. 110
     EPIC: Efficient Position-Independent Context Caching for Serving Large Language Models (ICML 2025)

     Large Language Model (LLM) · LLM Inference / Position-Independent Caching (PIC) · 2025
155. 111
     DistServe: Disaggregating Prefill and Decoding for Goodput-optimized Large Language Model Serving

     Large Language Model (LLM) · LLM Inference / Prefill-Decode (PD) Disaggregation · 2024
156. 112
     Inference without Interference: Disaggregate LLM Inference for Mixed Downstream Workloads (arXiv:2401.11181)

     Large Language Model (LLM) · LLM Inference / Prefill-Decode (PD) Disaggregation · 2024-01
157. 113
     Mooncake: A KVCache-centric Disaggregated Architecture for LLM Serving (FAST 2025)

     Large Language Model (LLM) · LLM Inference / Prefill-Decode (PD) Disaggregation · 2025
158. 114
     Prefill-as-a-Service: KVCache of Next-Generation Models Could Go Cross-Datacenter (arXiv:2604.15039)

     Large Language Model (LLM) · LLM Inference / Prefill-Decode (PD) Disaggregation · 2026-04
159. 115
     Splitwise: Efficient Generative LLM Inference Using Phase Splitting

     Large Language Model (LLM) · LLM Inference / Prefill-Decode (PD) Disaggregation · 2024
160. 116
     FastServe: Iteration-Level Preemptive Scheduling for Large Language Model Inference

     Large Language Model (LLM) · LLM Inference / Request Scheduling · 2026
161. 117
     Llumnix: Dynamic Scheduling for Large Language Model Serving

     Large Language Model (LLM) · LLM Inference / Request Scheduling · 2024
162. 118
     Orca: A Distributed Serving System for Transformer-Based Generative Models

     Large Language Model (LLM) · LLM Inference / Request Scheduling · 2022
163. 119
     Cache-Craft: Managing Chunk-Caches for Efficient Retrieval-Augmented Generation (SIGMOD 2025)

     Large Language Model (LLM) · LLM Inference / Retrieval-Augmented Generation (RAG) · 2025
164. 120
     CacheFocus: Dynamic Cache Re-Positioning for Efficient Retrieval-Augmented Generation (arXiv:2502.11101)

     Large Language Model (LLM) · LLM Inference / Retrieval-Augmented Generation (RAG) · 2025-02
165. 121
     RAGCache: Efficient Knowledge Caching for Retrieval-Augmented Generation (arXiv:2404.12457)

     Large Language Model (LLM) · LLM Inference / Retrieval-Augmented Generation (RAG) · 2024-04
166. 122
     FaaScale: Unlocking Fast LLM Scaling for Serverless Inference

     Large Language Model (LLM) · LLM Inference / Serverless Inference · 2026
167. 123
     HydraServe: Minimizing Cold Start Latency for Serverless LLM Serving in Public Clouds

     Large Language Model (LLM) · LLM Inference / Serverless Inference · 2026
168. 124
     ServerlessLLM: Low-Latency Serverless Inference for Large Language Models

     Large Language Model (LLM) · LLM Inference / Serverless Inference · 2024
169. 125
     Deja Vu: Contextual Sparsity for Efficient LLMs at Inference Time

     Large Language Model (LLM) · LLM Inference / Sparsity · 2023
170. 126
     InfiniGen: Efficient Generative Inference of Large Language Models with Dynamic KV Cache Management

     Large Language Model (LLM) · LLM Inference / Sparsity · 2024
171. 127
     PowerInfer: Fast Large Language Model Serving with a Consumer-grade GPU

     Large Language Model (LLM) · LLM Inference / Sparsity · 2024
172. 128
     Fast Inference from Transformers via Speculative Decoding

     Large Language Model (LLM) · LLM Inference / Speculative Decoding · 2023
173. 129
     Online Speculative Decoding

     Large Language Model (LLM) · LLM Inference / Speculative Decoding · 2024
174. 130
     SpecInfer: Accelerating Generative LLM Serving with Speculative Inference and Token Tree Verification

     Large Language Model (LLM) · LLM Inference / Speculative Decoding · 2024
175. 131
     Speculative Decoding with Big Little Decoder (NeurIPS 2023)

     Large Language Model (LLM) · LLM Inference / Speculative Decoding · 2023
176. 132
     Measuring Agents in Production

     Large Language Model (LLM) · LLM Inference / Workload Characterization · 2026
177. 133
     TraceLab: Characterizing Coding Agent Workloads for LLM Serving (arXiv:2606.30560)

     Large Language Model (LLM) · LLM Inference / Workload Characterization · 2026-06-25
178. 134
     Accelerating the Training of Large Language Models using Efficient Activation Rematerialization and Optimal Hybrid Parallelism

     Large Language Model (LLM) · LLM Training / Hybrid Parallelism · 2024
179. 135
     Alpa: Automating Inter- and Intra-Operator Parallelism for Distributed Deep Learning

     Large Language Model (LLM) · LLM Training / Hybrid Parallelism · 2022
180. 136
     Beat the long tail: Distribution-Aware Speculative Decoding for RL Training

     Large Language Model (LLM) · LLM Training / RL Post-Training · 2026
181. 137
     DynaRL: Flexible and Dynamic Scheduling of Large-Scale Reinforcement Learning Training

     Large Language Model (LLM) · LLM Training / RL Post-Training · 2026
182. 138
     HetRL: Efficient Reinforcement Learning for LLMs in Heterogeneous Environments

     Large Language Model (LLM) · LLM Training / RL Post-Training · 2026
183. 139
     History Doesn't Repeat Itself but Rollouts Rhyme: Accelerating Reinforcement Learning with RhymeRL

     Large Language Model (LLM) · LLM Training / RL Post-Training · 2026
184. 140
     RLinf: Flexible and Efficient Large-Scale Reinforcement Learning via Macro-to-Micro Flow Transformation

     Large Language Model (LLM) · LLM Training / RL Post-Training · 2026
185. 141
     ReSpec: Towards Optimizing Speculative Decoding in Reinforcement Learning Systems

     Large Language Model (LLM) · LLM Training / RL Post-Training · 2026
186. 142
     ReaL: Efficient RLHF Training of Large Language Models with Parameter Reallocation

     Large Language Model (LLM) · LLM Training / RL Post-Training · 2025
187. 143
     RobustRL: Role-Based Fault Tolerance System for RL Post-Training

     Large Language Model (LLM) · LLM Training / RL Post-Training · 2026
188. 144
     RollArt: Disaggregated Multi-Task Agentic RL Training at Scale

     Large Language Model (LLM) · LLM Training / RL Post-Training · 2026
189. 145
     RollPacker: Taming Long-Tail Rollouts for RL Post-Training with Tail Batching

     Large Language Model (LLM) · LLM Training / RL Post-Training · 2026
190. 146
     Seer: Online Context Learning for Fast Synchronous LLM Reinforcement Learning

     Large Language Model (LLM) · LLM Training / RL Post-Training · 2026
191. 147
     Taming the Long-Tail: Efficient Reasoning RL Training with Adaptive Drafter

     Large Language Model (LLM) · LLM Training / RL Post-Training · 2026
192. 148
     Weave: Efficient Co-Scheduling for Disaggregated RL Post-Training

     Large Language Model (LLM) · LLM Training / RL Post-Training · 2026
193. 149
     Bamboo: Making Preemptible Instances Resilient for Affordable Training of Large DNNs

     Large Language Model (LLM) · LLM Training / Reliability and Fault Tolerance · 2023
194. 150
     Gemini: Fast Failure Recovery in Distributed Training with In-Memory Checkpoints

     Large Language Model (LLM) · LLM Training / Reliability and Fault Tolerance · 2023
195. 151
     Holmes: Localizing Irregularities in LLM Training with Mega-scale GPU Clusters

     Large Language Model (LLM) · LLM Training / Reliability and Fault Tolerance · 2025
196. 152
     Large-Scale AI Infra Reliability: Challenges, Strategies, and Llama 3 Training Experience (DSN-S 2025)

     Large Language Model (LLM) · LLM Training / Reliability and Fault Tolerance · 2025
197. 153
     MegaScale-Omni: A Hyper-Scale, Workload-Resilient System for MultiModal LLM Training in Production

     Large Language Model (LLM) · LLM Training / Reliability and Fault Tolerance · 2026
198. 154
     MegaScale: Scaling Large Language Model Training to More Than 10,000 GPUs

     Large Language Model (LLM) · LLM Training / Reliability and Fault Tolerance · 2024
199. 155
     Oobleck: Resilient Distributed Training of Large Models Using Pipeline Templates

     Large Language Model (LLM) · LLM Training / Reliability and Fault Tolerance · 2023
200. 156
     Robust LLM Training Infrastructure at ByteDance

     Large Language Model (LLM) · LLM Training / Reliability and Fault Tolerance · 2025
201. 157
     SDCs in the Wild: Characterizing and Diagnosing SDC-Defective GPUs in Production LLM Training

     Large Language Model (LLM) · LLM Training / Reliability and Fault Tolerance · 2026
202. 158
     Safeguarding LLM Training at Scale: Online SDC Detection and Insights from 35 Million GPU Hours

     Large Language Model (LLM) · LLM Training / Reliability and Fault Tolerance · 2026
203. 159
     Morphling: Fast, Near-Optimal Auto-Configuration for Cloud-Native Model Serving

     Model Serving · Auto-Configuration for Model Serving · 2021
204. 160
     Serving Unseen Deep Learning Models with Near-Optimal Configurations: a Fast Adaptive Search Approach

     Model Serving · Auto-Configuration for Model Serving · 2022
205. 161
     Clipper: A Low-Latency Online Prediction Serving System

     Model Serving · Model Serving Systems · 2017
206. 162
     INFaaS: Automated Model-less Inference Serving

     Model Serving · Model Serving Systems · 2021
207. 163
     Microsecond-scale Preemption for Concurrent GPU-accelerated DNN Inferences

     Model Serving · Model Serving Systems · 2022
208. 164
     Paella: Low-latency Model Serving with Software-defined GPU Scheduling

     Model Serving · Model Serving Systems · 2023
209. 165
     TensorFlow-Serving: Flexible, High-Performance ML Serving (NIPS 2017 Workshop on ML Systems)

     Model Serving · Model Serving Systems · 2017
210. 166
     Usher: Holistic Interference Avoidance for Resource Optimized ML Inference

     Model Serving · Model Serving Systems · 2024
211. 167
     A Survey of Large-Scale Deep Learning Serving System Optimization: Challenges and Opportunities (arXiv 2111.14247)

     Model Serving · Survey · 2021-11
212. 168
     A Survey of Multi-Tenant Deep Learning Inference on GPU (MLSys 2022 Workshop on Cloud Intelligence / AIOps)

     Model Serving · Survey · 2022
213. 169
     Accelerating Distributed MoE Training and Inference with Lina

     Mixture of Experts (MoE) · MoE Inference · 2023
214. 170
     Optimizing Dynamic Neural Networks with Brainstorm

     Mixture of Experts (MoE) · MoE Inference · 2023
215. 171
     Accelerating Distributed MoE Training and Inference with Lina

     Mixture of Experts (MoE) · MoE Training · 2023
216. 172
     Janus: A Unified Distributed Training Framework for Sparse Mixture-of-Experts Models

     Mixture of Experts (MoE) · MoE Training · 2023
217. 173
     SmartMoE: Efficiently Training Sparsely-Activated Models through Combining Offline and Online Parallelization

     Mixture of Experts (MoE) · MoE Training · 2023
218. 174
     AlloX: Compute Allocation in Hybrid Clusters

     Resource Scheduler · Scheduling for DL Training Workloads · 2020
219. 175
     AntMan: Dynamic Scaling on GPU Clusters for Deep Learning

     Resource Scheduler · Scheduling for DL Training Workloads · 2020
220. 176
     Astraea: A Fair Deep Learning Scheduler for Multi-Tenant GPU Clusters (TPDS 2021)

     Resource Scheduler · Scheduling for DL Training Workloads · 2021
221. 177
     Balancing Efficiency and Fairness in Heterogeneous GPU Clusters for Deep Learning

     Resource Scheduler · Scheduling for DL Training Workloads · 2020
222. 178
     Blox: A Modular Toolkit for Deep Learning Schedulers

     Resource Scheduler · Scheduling for DL Training Workloads · 2024
223. 179
     CASSINI: Network-Aware Job Scheduling in Machine Learning Clusters

     Resource Scheduler · Scheduling for DL Training Workloads · 2024
224. 180
     Gandiva: Introspective Cluster Scheduling for Deep Learning

     Resource Scheduler · Scheduling for DL Training Workloads · 2018
225. 181
     Heterogeneity-Aware Cluster Scheduling Policies for Deep Learning Workloads

     Resource Scheduler · Scheduling for DL Training Workloads · 2020
226. 182
     HiveD: Sharing a GPU Cluster for Deep Learning with Guarantees

     Resource Scheduler · Scheduling for DL Training Workloads · 2020
227. 183
     Interference-aware Multiplexing for Deep Learning in GPU Clusters: A Middleware Approach

     Resource Scheduler · Scheduling for DL Training Workloads · 2023
228. 184
     Looking Beyond GPUs for DNN Scheduling on Multi-Tenant Clusters

     Resource Scheduler · Scheduling for DL Training Workloads · 2022
229. 185
     Lucid: A Non-intrusive, Scalable and Interpretable Scheduler for Deep Learning Training Jobs

     Resource Scheduler · Scheduling for DL Training Workloads · 2023
230. 186
     Lyra: Elastic Scheduling for Deep Learning Clusters

     Resource Scheduler · Scheduling for DL Training Workloads · 2023
231. 187
     MAPA: Multi-Accelerator Pattern Allocation Policy for Multi-Tenant GPU Servers (SC 2021)

     Resource Scheduler · Scheduling for DL Training Workloads · 2021
232. 188
     Multi-Resource Interleaving for Deep Learning Training

     Resource Scheduler · Scheduling for DL Training Workloads · 2022
233. 189
     Optimus: An Efficient Dynamic Resource Scheduler for Deep Learning Clusters (EuroSys 2018)

     Resource Scheduler · Scheduling for DL Training Workloads · 2018
234. 190
     Pollux: Co-adaptive Cluster Scheduling for Goodput-Optimized Deep Learning

     Resource Scheduler · Scheduling for DL Training Workloads · 2021
235. 191
     Shockwave: Fair and Efficient Cluster Scheduling for Dynamic Adaptation in Machine Learning

     Resource Scheduler · Scheduling for DL Training Workloads · 2023
236. 192
     Sia: Heterogeneity-aware, goodput-optimized ML-cluster scheduling

     Resource Scheduler · Scheduling for DL Training Workloads · 2023
237. 193
     Singularity: Planet-Scale, Preemptive and Elastic Scheduling of AI Workloads (arXiv 2202.07848)

     Resource Scheduler · Scheduling for DL Training Workloads · 2022
238. 194
     Themis: Fair and Efficient GPU Cluster Scheduling (NSDI 2020)

     Resource Scheduler · Scheduling for DL Training Workloads · 2020
239. 195
     Tiresias: A GPU Cluster Manager for Distributed Deep Learning

     Resource Scheduler · Scheduling for DL Training Workloads · 2019
240. 196
     Topology-Aware GPU Scheduling for Learning Workloads in Cloud Environments (SC 2017)

     Resource Scheduler · Scheduling for DL Training Workloads · 2017
241. 197
     SLAQ: Quality-Driven Scheduling for Distributed Machine Learning

     Resource Scheduler · Scheduling for General ML Training Workloads · 2017
242. 198
     Deep Learning Workload Scheduling in GPU Datacenters: Taxonomy, Challenges and Vision (arXiv 2205.11913)

     Resource Scheduler · Survey · 2022-05
243. 199
     Analysis of Large-Scale Multi-Tenant GPU Clusters for DNN Training Workloads (ATC 2019)

     Resource Scheduler · Trace Analysis · 2019
244. 200
     Characterization and Prediction of Deep Learning Workloads in Large-Scale GPU Datacenters (SC 2021)

     Resource Scheduler · Trace Analysis · 2021
245. 201
     Characterizing Deep Learning Training Workloads on Alibaba-PAI (IISWC 2019)

     Resource Scheduler · Trace Analysis · 2019
246. 202
     MLaaS in the Wild: Workload Analysis and Scheduling in Large-Scale Heterogeneous GPU Clusters

     Resource Scheduler · Trace Analysis · 2022
247. 203
     MSRL: Distributed Reinforcement Learning with Dataflow Fragments

     Reinforcement Learning (RL) · Uncategorized · 2023
248. 204
     Ray: A Distributed Framework for Emerging AI Applications

     Reinforcement Learning (RL) · Uncategorized · 2018
249. 2026
     [随笔]什么是AI Native

     2026-07-11 · 随笔/观点
250. 2026
     详细谈谈DSpark投机解码的原理

     2026-07-04 · 完整技术章节
251. 2026
     谈谈2026年ScaleUP标准的变化

     2026-06-26 · 完整系统章节
252. 2026
     用Agent来分析Nvidia RTX系列芯片

     2026-06-22 · 完整研究工作流章节
253. 2026
     谈谈如何使用Agentic Workflow进行大模型全栈研究

     2026-06-08 · 完整研究工作流章节
254. 2026
     谈谈 AWS 的随机图网络拓扑

     2026-06-03 · 完整网络系统章节
255. 2026
     使用Agent做Nvidia GPU芯片微架构分析

     2026-05-29 · 完整研究工作流章节
256. 2026
     [大模型中的数学-103] Grassmann流形视角下的Attention算法

     2026-05-25 · 待人工标注
257. 2026
     谈谈GFD vs cudaMemcpBatchAsync

     2026-05-25 · 完整技术章节
258. 2026
     [大模型中的数学-102] 范畴论视角下的Attention算法

     2026-05-24 · 待人工标注
259. 2026
     [大模型中的数学-101] 大语言模型中的“语言”: 是符号的幻影, 还是意义的载体?

     2026-05-24 · 待人工标注
260. 2026
     Agent写的一个高性能Host-to-Device 传输库

     2026-05-23 · 完整研究工作流章节
261. 2026
     DeepSeek-V4详细分析(2): MegaMoE

     2026-05-14 · 模型架构解读
262. 2026
     谈谈OpenAI 发布的MRC

     2026-05-07 · 待人工标注
263. 2026
     DeepEPv2分析(3)-EP Direct Dispatch/Combine Kernel

     2026-05-05 · 完整技术章节
264. 2026
     DeepEPv2分析(4)-EP Hybrid Dispatch Combine Kernel

     2026-05-05 · 完整技术章节
265. 2026
     DeepEPv2分析(1)

     2026-05-04 · 完整技术章节
266. 2026
     DeepEPv2分析(2)-EP Overview

     2026-05-04 · 完整技术章节
267. 2026
     DeepSeek-V4详细分析(1): 算法和模型结构

     2026-05-01 · 技术结构解读
268. 2026
     NCCL Gin & Symmetric Memory

     2026-05-01 · 完整技术章节
269. 2026
     谈谈Agentic时代的架构及互连

     2026-04-05 · 完整研究工作流章节
270. 2026
     Token交易所专题研究：AI能力商品化驱动新型交易所基础设施形成

     2026-03-22 · 待人工标注
271. 2026
     谈谈Kimi Attention Res的工作

     2026-03-21 · 技术机制解读
272. 2026
     Groq LPU架构详解

     2026-03-19 · 待人工标注
273. 2026
     详细谈谈Rubin + Groq 3 LPU架构

     2026-03-19 · 待人工标注
274. 2026
     AI需要新的数学语言?

     2026-03-16 · 待人工标注
275. 2026
     谈谈Feynman+LPU的架构

     2026-03-14 · 待人工标注
276. 2026
     谈谈DeepSeek DualPath以及ScaleOut直接接入存储

     2026-02-27 · 完整系统章节
277. 2026
     谈谈光互连的一些问题

     2026-02-26 · 完整系统章节
278. 2026
     谈谈Attention SInk及未来Attention算法设计

     2026-02-19 · 待人工标注
279. 2026
     从芯片架构的视角谈谈ScaleUP, smart or dumb

     2026-02-13 · 完整系统章节
280. 2026
     大模型推理系统(2)--体系结构的视角

     2026-02-09 · 完整系统章节
281. 2026
     从Moltbook社交到Agent Swarm

     2026-02-01 · 研究方法与协作讨论
282. 2026
     基于可训练先验的广义最优传输注意力机制

     2026-01-31 · 待人工标注
283. 2026
     再来谈谈Sparse Attention

     2026-01-26 · 技术机制解读
284. 2026
     谈谈基于ScaleUP的存储扩展和三网融合

     2026-01-25 · 待人工标注
285. 2026
     世界终究是属于草台班子的-2

     2026-01-23 · 待人工标注
286. 2026
     猜一猜DeepSeek-V4的模型架构

     2026-01-18 · 推测性分析
287. 2026
     “漫”谈RDMA现代化

     2026-01-16 · 待人工标注
288. 2026
     谈谈DeepSeek Engram: Conditional Memory

     2026-01-14 · 技术机制解读
289. 2026
     谈谈mHC-Lite:无需使用Sinkhorn-Knopp迭代的算法

     2026-01-13 · 待人工标注
290. 2026
     大语言模型推理硬件的挑战与研究方向

     2026-01-12 · 完整系统章节
291. 2026
     Nvidia如何整合吸收Groq的技术

     2026-01-07 · 待人工标注
292. 2026
     谈谈DeepSeek mHC

     2026-01-03 · 待人工标注
293. 2025
     谈谈那个被NV看上值20B的Groq

     2025-12-30 · 待人工标注
294. 2025
     谈谈字节的veRoCE, 但为啥不叫VeiWARP呢?

     2025-12-21 · 待人工标注
295. 2025
     谈谈AWS Trainium3/4 及未来的AI Infra演进

     2025-12-19 · 待人工标注
296. 2025
     旧文重发: 谈谈GPU云的经营风险和流动性管理

     2025-12-11 · 待人工标注
297. 2025
     谈谈Nested Learning

     2025-12-08 · 待人工标注
298. 2025
     谈谈Scale Across网络

     2025-12-04 · 待人工标注
299. 2025
     真的该看空NV么? 可能并不是!

     2025-11-26 · 待人工标注
300. 2025
     英伟达已经做的很好了? 吗?

     2025-11-22 · 待人工标注
301. 2025
     谈谈GPU微架构及系统架构的演进

     2025-11-11 · 待人工标注
302. 2025
     谈谈未来Attention算法的选择, Full, Sparse or Linear ?

     2025-11-08 · 技术比较
303. 2025
     Inside Nvidia GPU: 谈谈Blackwell的不足并预测一下Rubin的微架构

     2025-11-05 · 待人工标注
304. 2025
     Tensor-103.4 Blackwell GEMM

     2025-11-01 · 待人工标注
305. 2025
     Tensor-103.3 Hopper Persistent Kernel

     2025-10-27 · 待人工标注
306. 2025
     Tensor-103.2: Hopper GEMM

     2025-10-22 · 待人工标注
307. 2025
     Tensor-103.1: Basic GEMM

     2025-10-18 · 待人工标注
308. 2025
     谈谈ESUN, SUE和UALink

     2025-10-18 · 完整系统章节
309. 2025
     Tilelang-2: 基本操作

     2025-10-12 · 待人工标注
310. 2025
     Tensor-102: GEMV

     2025-10-09 · 待人工标注
311. 2025
     谈谈 Hierarchical Sparse Attention (HSA)

     2025-10-06 · 待人工标注
312. 2025
     Tensor-101: Element-Wise Add

     2025-10-05 · 待人工标注
313. 2025
     Tilelang-1: Introduction

     2025-10-02 · 待人工标注
314. 2025
     CuteDSL-2: 基本操作

     2025-09-30 · 待人工标注
315. 2025
     学习一下DeepSeek-V3.2

     2025-09-30 · 模型解读
316. 2025
     谈谈阿里的AI Infra磐久128卡超节点

     2025-09-25 · 完整系统章节
317. 2025
     CuTe Layout代数-1: Overview

     2025-09-24 · 待人工标注
318. 2025
     CuteDSL-1: Introduction

     2025-09-21 · 待人工标注
319. 2025
     谈谈UB, UALink以及NV和Intel的合作

     2025-09-20 · 完整系统章节
320. 2025
     详细分析一下Nvidia Rubin CPX

     2025-09-17 · 待人工标注
321. 2025
     学习一下Linear Layout

     2025-09-15 · 待人工标注
322. 2025
     从GPU缓存的视角看芯片设计和互连

     2025-09-13 · 完整系统章节
323. 2025
     谈谈芯片设计的一些取舍

     2025-09-07 · 待人工标注
324. 2025
     Firefly: 数据中心内的高精度时钟同步

     2025-09-05 · 待人工标注
325. 2025
     谈谈Google Falcon的可靠传输论文并对比分析CIPU eRDMA

     2025-09-05 · 完整系统章节
326. 2025
     谈谈RDMA和ScaleUP的可靠传输

     2025-08-31 · 完整系统章节
327. 2025
     Hotchip-2025:Day2 AI芯片和光互连

     2025-08-27 · 完整系统章节
328. 2025
     Hotchip-2025: Day1网络篇

     2025-08-26 · 完整系统章节
329. 2025
     Hotchip-2025: Day0 Tutorials记录

     2025-08-25 · 待人工标注
330. 2025
     谈谈Transformer的一些演进: UT,MoD,MoR...

     2025-08-25 · 待人工标注
331. 2025
     谈谈AutoGLM 2.0以及背后的Agent Infra

     2025-08-22 · 完整研究工作流章节
332. 2025
     大模型时代的数学基础(9)- SDPA和最优传输, 强化学习及信息几何的联系

     2025-08-21 · 待人工标注
333. 2025
     再来谈谈UltraEthernet的设计原则

     2025-08-17 · 完整系统章节
334. 2025
     谈谈搜广推场景的RDMA

     2025-08-11 · 待人工标注
335. 2025
     从GPT-OSS谈谈大模型算法和Infra演进

     2025-08-07 · 独立系统解释
336. 2025
     从一个基金经理的视角谈谈架构师的素养

     2025-08-06 · 待人工标注
337. 2025
     再来谈谈AFD

     2025-08-05 · 待人工标注
338. 2025
     从算法-系统-硬件协同的视角谈谈AFD

     2025-08-03 · 待人工标注
339. 2025
     谈谈Agentic AI对Infra的需求

     2025-07-22 · 基础设施研究讨论
340. 2025
     同构视角下的异构计算: 谈谈AI的软硬件交付界面, GPGPU or DSA

     2025-07-19 · 完整系统章节
341. 2025
     全球首个A股金融智能体?

     2025-07-18 · 待人工标注
342. 2025
     谈谈Kimi K2, 第一个开源的1T参数模型

     2025-07-12 · 独立系统解释
343. 2025
     谈谈AWS GB200实例, 顺便谈谈各种自研NPU的困境

     2025-07-10 · 待人工标注
344. 2025
     谈谈GPU云的经营风险和流动性管理

     2025-07-05 · 待人工标注
345. 2025
     从ODCC夏季会议再来谈谈ScaleUP的一些进展

     2025-07-01 · 完整系统章节
346. 2025
     学习华为CloudMatrix 384上LLM推理

     2025-06-18 · 待人工标注
347. 2025
     关于3D Stacking DRAM GPU的一些想法

     2025-06-14 · 完整系统章节
348. 2025
     谈谈SDR-RDMA, 所谓的软件定义的可靠性

     2025-06-09 · 完整系统章节
349. 2025
     从AI落地的视角看看Infra的需求

     2025-06-08 · 待人工标注
350. 2025
     再来谈谈ScaleUP和ScaleOut

     2025-05-31 · 完整系统章节
351. 2025
     锐评某友商说传统云还在卖铁: 从金融的视角谈云计算及其流动性管理

     2025-05-28 · 待人工标注
352. 2025
     一个关于AI编程/Agent的讨论

     2025-05-25 · 研究方法与编程讨论
353. 2025
     谈谈NV在ComputeX Keynote

     2025-05-20 · 待人工标注
354. 2025
     谈谈DeepSeek对AI架构硬件的思考

     2025-05-16 · 待人工标注
355. 2025
     CUDA Green Context

     2025-05-07 · 完整系统章节
356. 2025
     现代NVidia GPU架构

     2025-05-05 · 待人工标注
357. 2025
     民科: 预测一下未来五年大模型的架构?

     2025-05-03 · 待人工标注
358. 2025
     谈谈以太网GPU Scale-UP的工作EthLink

     2025-04-30 · 完整系统章节
359. 2025
     谈谈大模型可解释性(2)

     2025-04-30 · 待人工标注
360. 2025
     Test-Time ScalingLaw又失效了? 算力约束下的国产模型突破在何方?

     2025-04-27 · 待人工标注
361. 2025
     高带宽处理器, 又在基于PNM卷访存密集型算子?

     2025-04-26 · 待人工标注
362. 2025
     CUDA-Next: 基于任务的张量计算的DSL?

     2025-04-26 · 待人工标注
363. 2025
     谈谈腾讯DeepEP的优化

     2025-04-24 · 待人工标注
364. 2025
     再来谈谈GPU体系结构及互联

     2025-04-20 · 完整系统章节
365. 2025
     从Sutton苦涩的教训来谈谈AI基础设施

     2025-04-15 · 待人工标注
366. 2025
     谈谈GPU的内存模型及互联网络设计

     2025-04-13 · 完整系统章节
367. 2025
     谈谈AI时代的CIPU 2.0

     2025-04-10 · 完整系统章节
368. 2025
     谈谈Llama4 和DeepSeek GRM

     2025-04-06 · 待人工标注
369. 2025
     谈谈字节的Attention/Expert分离

     2025-04-05 · 独立系统解释
370. 2025
     大模型推理系统(1)--先从分布式系统DEC VAXCluster谈起

     2025-04-02 · 完整系统章节
371. 2025
     从GTC25谈谈GPU互联(修正版)

     2025-03-31 · 完整系统章节
372. 2025
     从GTC25谈谈GPU互联

     2025-03-30 · 待人工标注
373. 2025
     从3FS性能谈谈数据密集型应用上云的挑战和机会

     2025-03-29 · 完整系统章节
374. 2025
     Tensor-011 Blackwell TensorCore

     2025-03-26 · 待人工标注
375. 2025
     从GTC25谈谈大模型架构的演进

     2025-03-23 · 待人工标注
376. 2025
     谈谈3FS的写性能优化的难点

     2025-03-21 · 完整系统章节
377. 2025
     DeepSeek-V3/R1推理效率分析(2)

     2025-03-21 · 独立系统解释
378. 2025
     再以eRDMA的视角谈谈DeepEP和3FS

     2025-03-19 · 完整系统章节
379. 2025
     谈谈三万亿的破绽(6): 从GTC25上分析

     2025-03-19 · 待人工标注
380. 2025
     DeepSeek-V3/R1推理效率分析-Blackwell性能估计(V0.4)

     2025-03-18 · 独立系统解释
381. 2025
     DeepSeek 3FS解读与源码分析（4）：Meta Service解读

     2025-03-17 · 完整系统章节
382. 2025
     DeepSeek 3FS解读与源码分析（5）：客户端解读

     2025-03-17 · 完整系统章节
383. 2025
     DeepSeek-V3/R1推理效率分析(v0.17)

     2025-03-16 · 独立系统解释
384. 2025
     DeepSeek-V3/R1推理效率分析(v0.15)

     2025-03-15 · 独立系统解释
385. 2025
     DeepSeek-V3/R1推理效率分析

     2025-03-14 · 独立系统解释
386. 2025
     从DeepSeek MoE专家负载均衡谈起

     2025-03-10 · 待人工标注
387. 2025
     DeepSeek 3FS解读与源码分析（3）：Storage模块解读

     2025-03-09 · 完整系统章节
388. 2025
     谈谈三万亿的破绽(5): 以基金经理的视角

     2025-03-07 · 待人工标注
389. 2025
     RDMA这十年的反思4: 从DeepSeek的3FS和DeepEP谈起

     2025-03-07 · 完整系统章节
390. 2025
     谈谈字节的COMET, 另一个细粒度的MoE通信和计算Overlap方案

     2025-03-06 · 独立系统解释
391. 2025
     DeepSeek 3FS解读与源码分析（2）：网络通信模块分析

     2025-03-02 · 完整系统章节
392. 2025
     DeepSeek 3FS解读与源码分析（1）：高效训练之道

     2025-03-02 · 完整系统章节
393. 2025
     谈谈DeepSeek原厂推理方案

     2025-03-02 · 待人工标注
394. 2025
     基于eRDMA实测DeepSeek开源的3FS

     2025-03-01 · 完整系统章节
395. 2025
     [一个脑洞]MoE专家权重是否可以预测并Prefetch?

     2025-02-28 · 待人工标注
396. 2025
     分析一下EP并行和DeepSeek开源的DeepEP代码

     2025-02-27 · 独立系统解释
397. 2025
     谈谈Kimi的LLM相关的工作, K1.5/MOBA/MUON

     2025-02-23 · 待人工标注
398. 2025
     基于强化学习的模型商品化推荐系统及MaaS盈利模式分析

     2025-02-22 · 待人工标注
399. 2025
     从Deepseek R1和NSA算法谈谈个人的一些反思

     2025-02-19 · 待人工标注
400. 2025
     谈谈DeepSeek Native Sparse Attention

     2025-02-18 · 待人工标注
401. 2025
     谈谈微信+DeepSeek

     2025-02-16 · 待人工标注
402. 2025
     谈谈DeepSeek MoE模型优化和未来演进以及字节Ultra-Sparse Memory相关的工作

     2025-02-15 · 待人工标注
403. 2025
     详细谈谈DeepSeek MoE相关的技术发展

     2025-02-13 · 待人工标注
404. 2025
     谈谈DeepSeek-R1满血版推理部署和优化

     2025-02-13 · 待人工标注
405. 2025
     谈谈国产算力支持大模型和MoE/RL算法协同演进方向

     2025-02-13 · 待人工标注
406. 2025
     详细谈谈DeepSeek MoE相关的技术发展

     2025-02-10 · 待人工标注
407. 2025
     谈谈国产算力支持大模型和MoE/RL算法协同演进方向

     2025-02-06 · 待人工标注
408. 2025
     AI的中国资本叙事

     2025-01-31 · 待人工标注
409. 2025
     [去年旧文重发]谈谈AISys架构师的基本素养

     2025-01-28 · 待人工标注
410. 2025
     关于Blackwell架构的一些更正

     2025-01-27 · 待人工标注
411. 2025
     谈谈三万亿的破绽(4)

     2025-01-27 · 待人工标注
412. 2025
     Blackwell TensorCore架构

     2025-01-26 · 待人工标注
413. 2025
     一些关于Blackwell微架构的分析

     2025-01-25 · 待人工标注
414. 2025
     谈谈大模型架构的演进之路, The Art of memory.

     2025-01-23 · 待人工标注
415. 2025
     一个关于MoE的猜想

     2025-01-14 · 待人工标注
416. 2025
     Pretrain ScalingLaw真的终结了么?

     2025-01-05 · 待人工标注
417. 2025
     新年: 但愿那月落重生灯再红

     2025-01-01 · 待人工标注
418. 2025
     谈谈一个新的MoA模型架构DeepSick-4.1T

     2025-04-01 · 待人工标注

当前分类里没有匹配条目。

Systems for ML 的题名、分类和外链来自 Lingyun Yang 维护的公开目录快照；公众号部分只保存目录元数据，不转载正文。收录规则见证据与来源。