Skip to yearly menu bar Skip to main content


Show Detail
Schedule Mon Tue Wed Thu Fri
Timezone: US/Pacific
 
Filter Rooms:  

MON 18 MAY
8 a.m.
Registration
(ends 5:00 PM)
8:30 a.m.
Remarks:
Opening Remarks
(ends 8:45 AM)
8:45 a.m.
Oral:
Beyond Model Serving: Cross-Stack Co-Design for Agentic Systems
(ends 9:05 AM)
9:05 a.m.
Oral:
LMCache: An Efficient KV Cache Layer for Enterprise-Scale LLM Inference
(ends 9:20 AM)
9:20 a.m.
Oral:
Eliciting Language Model Behaviors with Investigator Agents
(ends 9:40 AM)
10 a.m.
Break:
Coffee Break
(ends 10:30 AM)
10:30 a.m.
Sponsor Lightning Talks 1 [10:30-11:30]
(ends 11:30 AM)
11:30 a.m.
Break:
Lunch Break
(ends 1:00 PM)
2 p.m.
Sponsor Lightning Talks 2 [2:00-3:00]
(ends 3:00 PM)
3 p.m.
Break:
Break
(ends 3:30 PM)
3:30 p.m.
Keynote Talk:
When AI Starts Writing Systems Code
Mark Saroufim
(ends 4:30 PM)
4:30 p.m.
YPS Poster Session & YPS Reception [4:30-6:30]
(ends 6:30 PM)

TUE 19 MAY
8 a.m.
Registration
(ends 5:00 PM)
8:30 a.m.
Remarks:
Opening Remarks
(ends 8:45 AM)
8:45 a.m.
Best Paper Session [8:45-10:00]
(ends 10:00 AM)
Overflow / Viewing [8:45-10:00]
(ends 10:00 AM)
10 a.m.
Break:
Break
(ends 10:30 AM)
10:30 a.m.
Keynote Talk:
The Next Horizon of Systems: From MLSys to System Intelligence
Lidong Zhou
(ends 11:30 AM)
Overflow / Viewing [10:30-11:30]
(ends 11:30 AM)
11:30 a.m.
Break:
Lunch Break
(ends 1:00 PM)
1 p.m.
Research-Track Oral Presentation: R16: Agentic AI [1:00-2:30]
Matrix: Peer-to-Peer Multi-Agent Synthetic Data Generation Framework
Optimizing PyTorch Inference with LLM-Based Multi-Agent Systems
OSWorld-Human: Benchmarking the Efficiency of Computer-Use Agents
VeriMoA: A Mixture-of-Agents Framework for Spec-to-HDL Generation
When Enough is Enough: Rank-Aware Early Termination for Vector Search
(ends 2:30 PM)
Research-Track Oral Presentation: R19: Multimodal and Generative Models [1:00-2:30]
db-SP: Accelerating Sparse Attention for Visual Generative Models with Dual-Balanced Sequence Parallelism
Flash3DGS: Algorithm and System Co-Optimization for Fast 3D Gaussian Splatting on GPUs
StreamDiffusionV2: A Streaming System for Dynamic and Interactive Video Generation
TiDAR: Think in Diffusion, Talk in Autoregression
TriInfer: Hybrid EPD Disaggregation for Efficient Multimodal Large Language Model Inference
(ends 2:30 PM)
2:30 p.m.
Break:
Break
(ends 2:45 PM)
2:45 p.m.
Research-Track Oral Presentation: R1: LLM Serving [2:45-4:15]
From Tokens to Layers: Redefining Stall-Free Scheduling for LLM Serving with Layered Prefill
HELIOS : Adaptive Model And Early-Exit Selection for Efficient LLM Inference Serving
PLA-Serve: A Prefill-Length-Aware LLM Serving System
Stream2LLM: Overlap Context Streaming and Prefill for Reduced Time-to-First-Token
SuperInfer: SLO-Aware Rotary Scheduling and Memory Management for LLM Inference on Superchips
TokenBlend: Accelerating Tensor Parallelism LLM Inference Through Efficient Compute-Communication Overlap
(ends 4:15 PM)
Research-Track Oral Presentation: R9: Agentic AI [2:45-4:15]
AgenticCache: Cache-Driven Asynchronous Planning for Embodied AI Agents
FlashAgents: Accelerating Multi-Agent LLM Systems via Streaming Prefill Overlap
Hippocampus: An Efficient and Scalable Memory Module for Agentic AI
Ontology-Guided Long-Term Memory for Conversational RAG
RAGBoost: Efficient Retrieval-Augmented Generation with Accuracy-Preserving Context Reuse
RagInfer: Efficient Retrieval-Augmented Generation Inference with Lookahead Retrieval
(ends 4:15 PM)
4:15 p.m.
Break:
Break
(ends 4:30 PM)
4:30 p.m.
Research-Track Oral Presentation: R8: Federated Learning [4:30-6:00]
DisAgg: Distributed Aggregators for Efficient Secure Aggregation
FLoRIST: Singular Value Thresholding for Efficient and Accurate Federated Fine-Tuning of Large Language Models
PLayer-FL: A Principled Approach to Personalized Layer-wise Cross-Silo Federated Learning
ProToken: Token-Level Attribution for Federated Large Language Models
SONAR: Benchmarking Topology and Collaboration in Decentralized Learning
Zero redundancy distributed learning with differential privacy
(ends 6:00 PM)
Research-Track Oral Presentation: R10: Security and Privacy [4:30-6:00]
Blueprint, Bootstrap, and Bridge: A Security Look at NVIDIA GPU Confidential Computing
CSLE: A Reinforcement Learning Platform for Autonomous Security Management
G-HEMP: FAST MULTI-GPU PRIVATE INFERENCE FOR LARGE-SCALE GCNS WITH HOMOMORPHIC ENCRYPTION
Privatar: Scalable Privacy-preserving Multi-user VR via Secure Offloading
Toward Principled LLM Safety Testing: Solving the Jailbreak Oracle Problem
ZK-APEX: ZERO-KNOWLEDGE APPROXIMATE PERSONALIZED UNLEARNING WITH EXECUTABLE PROOFS
(ends 6:00 PM)
6 p.m.
Poster Session 1 & Opening Reception [6:00-8:00]
Posters -
AgenticCache: Cache-Driven Asynchronous Planning for Embodied AI Agents
Blueprint, Bootstrap, and Bridge: A Security Look at NVIDIA GPU Confidential Computing
CSLE: A Reinforcement Learning Platform for Autonomous Security Management
db-SP: Accelerating Sparse Attention for Visual Generative Models with Dual-Balanced Sequence Parallelism
DisAgg: Distributed Aggregators for Efficient Secure Aggregation
Flash3DGS: Algorithm and System Co-Optimization for Fast 3D Gaussian Splatting on GPUs
FlashAgents: Accelerating Multi-Agent LLM Systems via Streaming Prefill Overlap
FLoRIST: Singular Value Thresholding for Efficient and Accurate Federated Fine-Tuning of Large Language Models
From Tokens to Layers: Redefining Stall-Free Scheduling for LLM Serving with Layered Prefill
G-HEMP: FAST MULTI-GPU PRIVATE INFERENCE FOR LARGE-SCALE GCNS WITH HOMOMORPHIC ENCRYPTION
HELIOS : Adaptive Model And Early-Exit Selection for Efficient LLM Inference Serving
Hippocampus: An Efficient and Scalable Memory Module for Agentic AI
Matrix: Peer-to-Peer Multi-Agent Synthetic Data Generation Framework
Ontology-Guided Long-Term Memory for Conversational RAG
Optimizing PyTorch Inference with LLM-Based Multi-Agent Systems
OSWorld-Human: Benchmarking the Efficiency of Computer-Use Agents
PLA-Serve: A Prefill-Length-Aware LLM Serving System
PLayer-FL: A Principled Approach to Personalized Layer-wise Cross-Silo Federated Learning
Privatar: Scalable Privacy-preserving Multi-user VR via Secure Offloading
ProToken: Token-Level Attribution for Federated Large Language Models
RAGBoost: Efficient Retrieval-Augmented Generation with Accuracy-Preserving Context Reuse
RagInfer: Efficient Retrieval-Augmented Generation Inference with Lookahead Retrieval
SONAR: Benchmarking Topology and Collaboration in Decentralized Learning
Stream2LLM: Overlap Context Streaming and Prefill for Reduced Time-to-First-Token
StreamDiffusionV2: A Streaming System for Dynamic and Interactive Video Generation
SuperInfer: SLO-Aware Rotary Scheduling and Memory Management for LLM Inference on Superchips
TiDAR: Think in Diffusion, Talk in Autoregression
TokenBlend: Accelerating Tensor Parallelism LLM Inference Through Efficient Compute-Communication Overlap
Toward Principled LLM Safety Testing: Solving the Jailbreak Oracle Problem
TriInfer: Hybrid EPD Disaggregation for Efficient Multimodal Large Language Model Inference
VeriMoA: A Mixture-of-Agents Framework for Spec-to-HDL Generation
When Enough is Enough: Rank-Aware Early Termination for Vector Search
Zero redundancy distributed learning with differential privacy
ZK-APEX: ZERO-KNOWLEDGE APPROXIMATE PERSONALIZED UNLEARNING WITH EXECUTABLE PROOFS
(ends 8:00 PM)

WED 20 MAY
8 a.m.
Registration
(ends 5:00 PM)
8:30 a.m.
Research-Track Oral Presentation: R6: LLM Training [8:30-10:00]
BOOST: BOttleneck-Optimized Scalable Training Framework for Low-Rank Large Language Models
Efficient Long-Context Language Model Training by Core Attention Disaggregation
MTraining: Distributed Dynamic Sparse Attention for Efficient Ultra-Long Context Training
NEST: Network- and Memory-Aware Device Placement for Distributed Deep Learning
ProTrain: Efficient LLM Training via Automatic Memory Management
Unleashing Scalable Context Parallelism for Foundation Models Pre-Training via FCP
(ends 10:00 AM)
Research-Track Oral Presentation: R2: LLM Serving [8:30-10:00]
BEAM: Joint Resource–Power Optimization for Energy-Efficient LLM Inference under SLO contraints
BOute: Cost-Efficient LLM Serving with Heterogeneous LLMs and GPUs via Multi-Objective Bayesian Optimization
Breaking the Ice: Analyzing Cold Start Latency in vLLM
FaaScale: Unlocking Fast LLM Scaling for Serverless Inference
Locality-Aware Beam Scheduling for Efficient Test-Time Compute with a Consumer-grade GPU
MorphServe: Efficient and Workload-Aware LLM Serving via Runtime Quantized Layer Swapping and KV Cache Resizing
(ends 10:00 AM)
10 a.m.
Break:
Break
(ends 10:30 AM)
10:30 a.m.
Overflow / Viewing [10:30-11:30]
(ends 11:30 AM)
Keynote Talk:
Keynote - Amin Vahdat
Amin Vahdat
(ends 11:30 AM)
11:30 a.m.
Break:
Lunch Break
(ends 1:00 PM)
1 p.m.
Research-Track Oral Presentation: R3: LLM Serving [1:00-2:30]
FlexiCache: Leveraging Temporal Stability of Attention Heads for Efficient KV Cache Management
Kitty: Accurate and Efficient 2-bit KV Cache Quantization with Dynamic Channel-wise Precision Boost
MAC-Attention: a Match--Amend--Complete scheme for fast and accurate attention computation
OPKV: A High-Throughput Plugin-Driven Framework for Recallable Sparsity in Paged KV Cache Systems
SkipKV: Selective Skipping of KV Generation and Storage for Efficient Inference with Large Reasoning Models
Using Span Queries to Optimize Cache and Attention Locality
(ends 2:30 PM)
Research-Track Oral Presentation: R11: ML for Systems [1:00-2:30]
Automated Algorithm Design for Auto-Tuning Optimizers
LLMInfer-Bench: Building the Virtuous Cycle for AI-driven LLM Systems
Practical Adversarial Multi-Armed Bandits with Sublinear Runtime
Unified LLM Model for Power, Performance, and Area Prediction from Hardware Code
Virtual Machine NUMA Placement at Scale: Learning the Norm, Shielding the Tail
When Machine Learning Isn’t Sure: Building Resilient ML-Based Computer Systems by Embracing Uncertainty
(ends 2:30 PM)
2:30 p.m.
Break:
Break
(ends 2:45 PM)
2:45 p.m.
Research-Track Oral Presentation: R5: LLM Serving [2:45-4:15]
CRAFT: Cost-aware Expert Replica Allocation with Fine-Grained Layerwise Estimations
Demystifying the Mixture of Experts Serving Tax
FarSkip-Collectives: Unhobbling Blocking Communication in Mixture of Experts Models
FP8-Flow-MoE: A Casting-Free FP8 Recipe without Double Quantization Error
GhostServe: A Lightweight Checkpointing System in the Shadow for Fault-Tolerant LLM Serving
RaidServe: High-performance Resilient Serving
(ends 4:15 PM)
Research-Track Oral Presentation: R12: LLM Training [2:45-4:15]
Beat the long tail: Distribution-Aware Speculative Decoding for RL Training
FlexTrain: Scalable Hybrid-Parallel Training with Elastic Resource Utilization and Consistent Accuracy
HetRL: Efficient Reinforcement Learning for LLMs in Heterogeneous Environments
Learning from Less: Measuring the Effectiveness of RLVR in Low Data and Compute Regimes
NexSpec: Towards Optimizing Speculative Decoding in Reinforcement Learning Systems
Pylo: Towards Accessible Learned Optimizers in PyTorch
(ends 4:15 PM)
4:15 p.m.
Break:
Break
(ends 4:30 PM)
4:30 p.m.
Research-Track Oral Presentation: R7: LLM Training [4:30-6:00]
A Lightweight High-Throughput Collective-Capable NoC for Large-Scale ML Accelerators
DreamDDP: Accelerating Low-Bandwidth Geo-Distributed LLM Training with Layer-wise Partial Synchronization
GriNNder: Breaking the Memory Capacity Wall in Full-Graph GNN Training with Storage Offloading
Grolar: Efficient LLM Training on Heterogeneous Clusters
HexiScale: Facilitating Large Language Model Training over Heterogeneous Hardware
RDMA Point-to-Point Communication for LLM Systems
(ends 6:00 PM)
Research-Track Oral Presentation: R15: Model Compression [4:30-6:00]
CAGE: Curvature-Aware Gradient Estimation For Accurate Quantization-Aware Training
MixLLM: LLM Quantization with Global Mixed-precision between Output-features and Highly-efficient System Design
Once-for-All Channel Mixers (HyperTinyPW): Generative Compression for TinyML
Search Your NVFP4 Scales!
Shannonic: Efficient Entropy-Optimal Compression for ML Workloads
(ends 6:00 PM)
6 p.m.
Poster Session 2 [6:00-8:00]
Posters -
A Lightweight High-Throughput Collective-Capable NoC for Large-Scale ML Accelerators
Automated Algorithm Design for Auto-Tuning Optimizers
BEAM: Joint Resource–Power Optimization for Energy-Efficient LLM Inference under SLO contraints
Beat the long tail: Distribution-Aware Speculative Decoding for RL Training
BOOST: BOttleneck-Optimized Scalable Training Framework for Low-Rank Large Language Models
BOute: Cost-Efficient LLM Serving with Heterogeneous LLMs and GPUs via Multi-Objective Bayesian Optimization
Breaking the Ice: Analyzing Cold Start Latency in vLLM
CAGE: Curvature-Aware Gradient Estimation For Accurate Quantization-Aware Training
CRAFT: Cost-aware Expert Replica Allocation with Fine-Grained Layerwise Estimations
Demystifying the Mixture of Experts Serving Tax
DreamDDP: Accelerating Low-Bandwidth Geo-Distributed LLM Training with Layer-wise Partial Synchronization
Efficient Long-Context Language Model Training by Core Attention Disaggregation
FaaScale: Unlocking Fast LLM Scaling for Serverless Inference
FarSkip-Collectives: Unhobbling Blocking Communication in Mixture of Experts Models
FlexiCache: Leveraging Temporal Stability of Attention Heads for Efficient KV Cache Management
FlexTrain: Scalable Hybrid-Parallel Training with Elastic Resource Utilization and Consistent Accuracy
FP8-Flow-MoE: A Casting-Free FP8 Recipe without Double Quantization Error
GhostServe: A Lightweight Checkpointing System in the Shadow for Fault-Tolerant LLM Serving
GriNNder: Breaking the Memory Capacity Wall in Full-Graph GNN Training with Storage Offloading
Grolar: Efficient LLM Training on Heterogeneous Clusters
HetRL: Efficient Reinforcement Learning for LLMs in Heterogeneous Environments
HexiScale: Facilitating Large Language Model Training over Heterogeneous Hardware
Kitty: Accurate and Efficient 2-bit KV Cache Quantization with Dynamic Channel-wise Precision Boost
Learning from Less: Measuring the Effectiveness of RLVR in Low Data and Compute Regimes
LLMInfer-Bench: Building the Virtuous Cycle for AI-driven LLM Systems
Locality-Aware Beam Scheduling for Efficient Test-Time Compute with a Consumer-grade GPU
MAC-Attention: a Match--Amend--Complete scheme for fast and accurate attention computation
MixLLM: LLM Quantization with Global Mixed-precision between Output-features and Highly-efficient System Design
MorphServe: Efficient and Workload-Aware LLM Serving via Runtime Quantized Layer Swapping and KV Cache Resizing
MTraining: Distributed Dynamic Sparse Attention for Efficient Ultra-Long Context Training
NEST: Network- and Memory-Aware Device Placement for Distributed Deep Learning
NexSpec: Towards Optimizing Speculative Decoding in Reinforcement Learning Systems
Once-for-All Channel Mixers (HyperTinyPW): Generative Compression for TinyML
OPKV: A High-Throughput Plugin-Driven Framework for Recallable Sparsity in Paged KV Cache Systems
Practical Adversarial Multi-Armed Bandits with Sublinear Runtime
ProTrain: Efficient LLM Training via Automatic Memory Management
Pylo: Towards Accessible Learned Optimizers in PyTorch
RaidServe: High-performance Resilient Serving
RDMA Point-to-Point Communication for LLM Systems
Search Your NVFP4 Scales!
Shannonic: Efficient Entropy-Optimal Compression for ML Workloads
SkipKV: Selective Skipping of KV Generation and Storage for Efficient Inference with Large Reasoning Models
Unified LLM Model for Power, Performance, and Area Prediction from Hardware Code
Unleashing Scalable Context Parallelism for Foundation Models Pre-Training via FCP
Using Span Queries to Optimize Cache and Attention Locality
Virtual Machine NUMA Placement at Scale: Learning the Norm, Shielding the Tail
When Machine Learning Isn’t Sure: Building Resilient ML-Based Computer Systems by Embracing Uncertainty
(ends 8:00 PM)

THU 21 MAY
8 a.m.
Registration
(ends 5:00 PM)
8:30 a.m.
Industry-Track Oral Presentation: I5: Agentic AI/MLSys [8:30-10:00]
ADS: AN AGENTIC DETECTION SYSTEM FOR ENTERPRISE AGENTIC AI SECURITY
Agentic Operator Generation for ML ASICs
Cost-aware Duration Prediction for Software Upgrades in Datacenters
PROMPTS: PeRformance Optimization via Multi-Agent Planning for LLM Training and Serving
The OpenHands Software Agent SDK: A Composable and Extensible Foundation for Production Agents
(ends 10:00 AM)
Research-Track Oral Presentation: R13: LLM Serving [8:30-10:00]
Accelerating Large-Scale Reasoning Model Inference with Sparse Self-Speculative Decoding
CDLM: CONSISTENCY DIFFUSION LANGUAGE MODELS FOR FASTER SAMPLING
PRISM: PARAMETRICALLY RESTRUCTURED INFERENCE FOR SPECULATIVE SAMPLING DRAFT MODELS
SpecDiff-2: Scaling Diffusion Drafter Alignment For Faster Speculative Decoding
Speculative Decoding: Performance or Illusion?
(ends 10:00 AM)
10 a.m.
Break:
Break
(ends 10:30 AM)
10:30 a.m.
Keynote Talk:
Keynote - Luke Zettlemoyer
Luke Zettlemoyer
(ends 11:30 AM)
Overflow / Viewing [10:30-11:30]
(ends 11:30 AM)
11:30 a.m.
Break:
Lunch Break
(ends 1:00 PM)
1 p.m.
Research-Track Oral Presentation: R17: Edge and Mobile [1:00-2:30]
EarthSight: A Distributed Framework for Low-Latency Satellite Intelligence
IntAttention: A Fully Integer Attention Pipeline for Efficient Edge Inference
LEANN: A Low-Storage Overhead Vector Index
REPARO: LOSS-RESILIENT GENERATIVE CODEC FOR VIDEO CONFERENCING
Rethinking DVFS for Mobile LLMs: Unified Energy-Aware Scheduling with CORE
(ends 2:30 PM)
Industry-Track Oral Presentation: I1: LLM Serving [1:00-2:30]
BatchLLM: Optimizing Large Batched LLM Inference with Global Prefix Sharing and Throughput-oriented Token Batching
Beyond the Buzz: A Pragmatic Take on Inference Disaggregation
Meeting SLOs, Slashing Hours: Automated Enterprise LLM Optimization with OptiKIT
Optimizing Deployment Configurations for LLM Inference
Scaling Up Large Language Models Serving Systems for Semantic Job Search
SHIP: SRAM-Based Huge Inference Pipelines for Fast LLM Serving
(ends 2:30 PM)
2:30 p.m.
Break:
Break
(ends 2:45 PM)
2:45 p.m.
Research-Track Oral Presentation: R18: Benchmarks [2:45-4:15]
Charon: A Unified and Fine-Grained Simulator for Large-Scale LLM Training and Inference
DriftBench: Measuring and Predicting Infrastructure Drift in LLM Serving Systems
Hawkeye: Reproducing GPU-Level Non-Determinism
Massive-Scale Out-Of-Core UMAP on the GPU
Parrot: Persuasion and Agreement Robustness Rating of Output Truth
(ends 4:15 PM)
Industry-Track Oral Presentation: I2: LLM Training [2:45-4:15]
AXLearn: Modular, Hardware-Agnostic Large Model Training
FlexScale: Flexible and High-Performance FSDP at Scale
FreeScale: Distributed Training for Sequence Recommendation Models with Minimal Scaling Cost
MoEBlaze: Breaking the Memory Wall for Efficient MoE Training on Modern GPUs
NodeSweep: Practical Straggler Detection and Health Monitoring for Large-Scale Foundation Model Training
Sparing Strategies to Minimize Reliability Impact On Large Training Jobs
(ends 4:15 PM)
4:15 p.m.
Break:
Break
(ends 4:30 PM)
4:30 p.m.
Research-Track Oral Presentation: R4: Compilers and Kernels [4:30-6:00]
AccelOpt: A Self-Improving LLM Agentic System for AI Accelerator Kernel Optimization
ApproxMLIR : Accuracy-Aware Compiler for Compound ML System
Event Tensor: A Unified Abstraction for Compiling Dynamic Megakernel
HipKittens: Fast and Furious AMD Kernels
ParallelKittens: Systematic and Practical Simplification of Multi-GPU AI Kernels
SchedFlow: Transparent and Flexible Intra-Device Parallelism via Programmable Operator Scheduling
(ends 6:00 PM)
Industry-Track Oral Presentation: I3: Benchmarks [4:30-6:00]
AIRS: Scaling Live Inference in Resource Constrained Environments
MLCommons Chakra: Advancing Performance Benchmarking and Co-design using Standardized Execution Traces
ML Fleet Efficiency: Improving TPU Systems at Scale with ML Productivity Goodput
ProfInfer: An eBPF-based Fine-Grained LLM Inference Profiler
SAKURAONE: An Open Ethernet–Based AI HPC System and Its Observed Workload Dynamics in a Single-Tenant LLM Development Environment
XProf: An Open, Scalable, and Extensible Profiling System for the Modern ML Stack
(ends 6:00 PM)
6 p.m.
Poster Session 3 [6:00-8:00]
Posters -
Accelerating Large-Scale Reasoning Model Inference with Sparse Self-Speculative Decoding
AccelOpt: A Self-Improving LLM Agentic System for AI Accelerator Kernel Optimization
ADS: AN AGENTIC DETECTION SYSTEM FOR ENTERPRISE AGENTIC AI SECURITY
Agentic Operator Generation for ML ASICs
AIRS: Scaling Live Inference in Resource Constrained Environments
ApproxMLIR : Accuracy-Aware Compiler for Compound ML System
Attribution-based Sparse Activation in Large Language Models
AXLearn: Modular, Hardware-Agnostic Large Model Training
BatchLLM: Optimizing Large Batched LLM Inference with Global Prefix Sharing and Throughput-oriented Token Batching
Beyond the Buzz: A Pragmatic Take on Inference Disaggregation
BLASST: Dynamic BLocked Attention Sparsity via Softmax Thresholding
CATWILD: Compiler Autotuning for TPU workloads in the Wild
CDLM: CONSISTENCY DIFFUSION LANGUAGE MODELS FOR FASTER SAMPLING
Charon: A Unified and Fine-Grained Simulator for Large-Scale LLM Training and Inference
Cost-aware Duration Prediction for Software Upgrades in Datacenters
Dataflow Is All You Need
DriftBench: Measuring and Predicting Infrastructure Drift in LLM Serving Systems
EarthSight: A Distributed Framework for Low-Latency Satellite Intelligence
Efficient, VRAM-Constrained xLM Inference on Clients
Event Tensor: A Unified Abstraction for Compiling Dynamic Megakernel
ExecuTorch - A Unified PyTorch Solution to Run ML Models On-Device
FlashAttention-4: Algorithm and Kernel Pipelining Co-Design for Asymmetric Hardware Scaling
Flashlight: PyTorch Compiler Extensions to Accelerate Attention Variants
FlexScale: Flexible and High-Performance FSDP at Scale
FreeScale: Distributed Training for Sequence Recommendation Models with Minimal Scaling Cost
Hawkeye: Reproducing GPU-Level Non-Determinism
HipKittens: Fast and Furious AMD Kernels
IntAttention: A Fully Integer Attention Pipeline for Efficient Edge Inference
LEANN: A Low-Storage Overhead Vector Index
Massive-Scale Out-Of-Core UMAP on the GPU
Meeting SLOs, Slashing Hours: Automated Enterprise LLM Optimization with OptiKIT
MLCommons Chakra: Advancing Performance Benchmarking and Co-design using Standardized Execution Traces
ML Fleet Efficiency: Improving TPU Systems at Scale with ML Productivity Goodput
MoEBlaze: Breaking the Memory Wall for Efficient MoE Training on Modern GPUs
NodeSweep: Practical Straggler Detection and Health Monitoring for Large-Scale Foundation Model Training
Optimizing Deployment Configurations for LLM Inference
ParallelKittens: Systematic and Practical Simplification of Multi-GPU AI Kernels
Parrot: Persuasion and Agreement Robustness Rating of Output Truth
PRISM: PARAMETRICALLY RESTRUCTURED INFERENCE FOR SPECULATIVE SAMPLING DRAFT MODELS
ProfInfer: An eBPF-based Fine-Grained LLM Inference Profiler
PROMPTS: PeRformance Optimization via Multi-Agent Planning for LLM Training and Serving
REPARO: LOSS-RESILIENT GENERATIVE CODEC FOR VIDEO CONFERENCING
Rethinking DVFS for Mobile LLMs: Unified Energy-Aware Scheduling with CORE
SAKURAONE: An Open Ethernet–Based AI HPC System and Its Observed Workload Dynamics in a Single-Tenant LLM Development Environment
Scaling Up Large Language Models Serving Systems for Semantic Job Search
SchedFlow: Transparent and Flexible Intra-Device Parallelism via Programmable Operator Scheduling
SHIP: SRAM-Based Huge Inference Pipelines for Fast LLM Serving
Sparing Strategies to Minimize Reliability Impact On Large Training Jobs
SpecDiff-2: Scaling Diffusion Drafter Alignment For Faster Speculative Decoding
Speculative Decoding: Performance or Illusion?
Spira: Exploiting Voxel Data Structural Properties for Efficient Sparse Convolution in Point Cloud Networks
The OpenHands Software Agent SDK: A Composable and Extensible Foundation for Production Agents
WAVE: A SYMBOLIC PYTHON DSL AND COMPILER FOR HIGH PERFORMANCE MACHINE LEARNING
XProf: An Open, Scalable, and Extensible Profiling System for the Modern ML Stack
(ends 8:00 PM)

FRI 22 MAY
8 a.m.
Registration
(ends 1:00 PM)
8:15 a.m.
Industry-Track Oral Presentation: I4: Compilers/HW [8:15-9:30]
CATWILD: Compiler Autotuning for TPU workloads in the Wild
Dataflow Is All You Need
Efficient, VRAM-Constrained xLM Inference on Clients
ExecuTorch - A Unified PyTorch Solution to Run ML Models On-Device
WAVE: A SYMBOLIC PYTHON DSL AND COMPILER FOR HIGH PERFORMANCE MACHINE LEARNING
(ends 9:30 AM)
Research-Track Oral Presentation: R14: Efficient Computation [8:15-9:30]
Attribution-based Sparse Activation in Large Language Models
BLASST: Dynamic BLocked Attention Sparsity via Softmax Thresholding
FlashAttention-4: Algorithm and Kernel Pipelining Co-Design for Asymmetric Hardware Scaling
Flashlight: PyTorch Compiler Extensions to Accelerate Attention Variants
Spira: Exploiting Voxel Data Structural Properties for Efficient Sparse Convolution in Point Cloud Networks
(ends 9:30 AM)
9:30 a.m.
Break:
Break
(ends 9:45 AM)
9:45 a.m.
Overflow / Viewing [9:45-10:45]
(ends 10:45 AM)
Keynote Talk:
Keynote - Christos Kozyrakis
Christos Kozyrakis
(ends 10:45 AM)
10:45 a.m.
Break:
Break
(ends 11:00 AM)
11 a.m.
Overflow / Viewing [11:00-1:00]
(ends 1:00 PM)
Competition:
Competition Track (Amazon Google NVIDIA)
(ends 1:00 PM)