HexAGenT reduces the SLO scale required for timely agentic LLM workflow completion by an average of 20.1% at 95% attainment and 33.0% at 99% attainment on heterogeneous A100/H100/H200 clusters.
hub
InProceedings of the 30th ACM International Conference on Architectural Support for Programming Languages and Operating Systems, Volume 1(Rotterdam, Netherlands)(ASPLOS ’25)
12 Pith papers cite this work. Polarity classification is still indexing.
12
Pith papers citing it
hub tools
citation-role summary
background 2
citation-polarity summary
roles
background 2polarities
background 2representative citing papers
ReaLB balances multimodal MoE inference loads by switching vision-heavy experts to lower FP4 precision per device rank, hiding the change in the dispatch phase to deliver 1.10-1.32x speedup with <1% accuracy degradation.
HybridGen achieves 1.41x-3.2x average speedups over six prior KV cache methods for LLM inference by using attention logit parallelism, a feedback-driven scheduler, and semantic-aware KV cache mapping.
A new abstract interpretation algorithm enables sound optimistic analysis of e-graphs during equality saturation, unifying it with non-destructive rewriting and improving precision on cyclic SSA programs.
ACALSim is a new simulation framework with customizable threading, event-driven execution, and shared-memory model that reports over 14x speedup versus SST and enables simulation of large LLaMA models that SST cannot complete.
NasZip delivers up to 8.4x speedup over CPU baselines and 1.69x over prior NDP accelerators for ANNS by combining near-data processing with statistics-based PCA early exiting, dynamic-float encoding, and data-aware neighbor mapping.
Empirical comparison of equality saturation versus stochastic search on five benchmarks to evaluate if e-graphs are superior for rewrite-based optimization.
ELMoE-3D achieves 6.6x average speedup and 4.4x energy efficiency gain for MoE serving on 3D hardware by scaling expert and bit elasticity for elastic self-speculative decoding.
AEGIS reduces inter-GPU communication by up to 81.3% in self-attention and reaches 96.62% scaling efficiency with 3.86x speedup on four GPUs for 2048-token encrypted Transformer inference.
CCCL delivers 1.34-1.94x faster cross-node GPU collectives via CXL memory pooling than 200 Gbps InfiniBand RDMA, with 1.11x LLM training speedup and 2.75x hardware cost reduction.
EnergAIzer predicts module-level GPU utilization from structured kernel patterns and feeds it into a power model to estimate dynamic power with 8% error on Ampere GPUs and 7% on H100 forecasts.
Aquas delivers a holistic hardware-software co-optimization framework on MLIR that models memory interfaces with cache effects and uses an e-graph retargetable compiler, achieving up to 15.61x speedup with 14.5% area overhead across four domains.
citing papers explorer
-
HexAGenT: Efficient Agentic LLM Serving via Workflow- and Heterogeneity-Aware Scheduling
HexAGenT reduces the SLO scale required for timely agentic LLM workflow completion by an average of 20.1% at 95% attainment and 33.0% at 99% attainment on heterogeneous A100/H100/H200 clusters.
-
ReaLB: Real-Time Load Balancing for Multimodal MoE Inference
ReaLB balances multimodal MoE inference loads by switching vision-heavy experts to lower FP4 precision per device rank, hiding the change in the dispatch phase to deliver 1.10-1.32x speedup with <1% accuracy degradation.
-
HybridGen: Efficient LLM Generative Inference via CPU-GPU Hybrid Computing
HybridGen achieves 1.41x-3.2x average speedups over six prior KV cache methods for LLM inference by using attention logit parallelism, a feedback-driven scheduler, and semantic-aware KV cache mapping.
-
Optimism in Equality Saturation
A new abstract interpretation algorithm enables sound optimistic analysis of e-graphs during equality saturation, unifying it with non-destructive rewriting and improving precision on cyclic SSA programs.
-
ACALSim: A Scalable Parallel Simulation Framework for High-Performance System Design Space Exploration
ACALSim is a new simulation framework with customizable threading, event-driven execution, and shared-memory model that reports over 14x speedup versus SST and enables simulation of large LLaMA models that SST cannot complete.
-
Rewrite System Showdown: Stochastic Search vs. EqSat
Empirical comparison of equality saturation versus stochastic search on five benchmarks to evaluate if e-graphs are superior for rewrite-based optimization.
-
ELMoE-3D: Leveraging Intrinsic Elasticity of MoE for Hybrid-Bonding-Enabled Self-Speculative Decoding in On-Premises Serving
ELMoE-3D achieves 6.6x average speedup and 4.4x energy efficiency gain for MoE serving on 3D hardware by scaling expert and bit elasticity for elastic self-speculative decoding.
-
AEGIS: Scaling Long-Sequence Homomorphic Encrypted Transformer Inference via Hybrid Parallelism on Multi-GPU Systems
AEGIS reduces inter-GPU communication by up to 81.3% in self-attention and reaches 96.62% scaling efficiency with 3.86x speedup on four GPUs for 2048-token encrypted Transformer inference.
-
CCCL: Node-Spanning GPU Collectives with CXL Memory Pooling
CCCL delivers 1.34-1.94x faster cross-node GPU collectives via CXL memory pooling than 200 Gbps InfiniBand RDMA, with 1.11x LLM training speedup and 2.75x hardware cost reduction.
-
EnergAIzer: Fast and Accurate GPU Power Estimation Framework for AI Workloads
EnergAIzer predicts module-level GPU utilization from structured kernel patterns and feeds it into a power model to estimate dynamic power with 8% error on Ampere GPUs and 7% on H100 forecasts.
-
Aquas: Enhancing Domain Specialization through Holistic Hardware-Software Co-Optimization based on MLIR
Aquas delivers a holistic hardware-software co-optimization framework on MLIR that models memory interfaces with cache effects and uses an e-graph retargetable compiler, achieving up to 15.61x speedup with 14.5% area overhead across four domains.