FlashAttention-3 achieves 1.5-2x speedup on H100 GPUs for attention, reaching 740 TFLOPs/s (75% utilization) in FP16 and near 1.2 PFLOPs/s in FP8 while cutting numerical error by 2.6x versus baseline FP8 attention.
Salmon, Mark A
5 Pith papers cite this work, alongside 154 external citations. Polarity classification is still indexing.
5
Pith papers citing it
154
external citations · external index
citation-role summary
method 1
citation-polarity summary
roles
method 1polarities
use method 1representative citing papers
TLX introduces MIMW-based extensions to Triton that let developers orchestrate warp-group execution and asynchronous hardware features while preserving blocked programming productivity, with kernels deployed in large-scale training and inference.
Diffusion Restore uses diffusion-based nonreversible local dynamics in the Restore MCMC framework for light transport, outperforming prior methods and achieving real-time GPU performance.
A trace-driven simulator models Hopper TMA/WGMMA pipelines at WarpGroup granularity, matching H800 FlashAttention-3 latency to 5.7% MAPE, and its analytical model explains GenZ's long-sequence DRAM underestimation.
RGoT uses RL to adaptively generate task-specific graphs of operations for GoT-style LLM prompting from a human-provided set, with results suggesting feasibility under constraints.
citing papers explorer
-
FlashAttention-3: Fast and Accurate Attention with Asynchrony and Low-precision
FlashAttention-3 achieves 1.5-2x speedup on H100 GPUs for attention, reaching 740 TFLOPs/s (75% utilization) in FP16 and near 1.2 PFLOPs/s in FP8 while cutting numerical error by 2.6x versus baseline FP8 attention.
-
TLX: Hardware-Native, Evolvable MIMW GPU Compiler for Large-scale Production Environments
TLX introduces MIMW-based extensions to Triton that let developers orchestrate warp-group execution and asynchronous hardware features while preserving blocked programming productivity, with kernels deployed in large-scale training and inference.
-
Diffusion Restore: Real-Time Markov Chain Monte Carlo Light Transport
Diffusion Restore uses diffusion-based nonreversible local dynamics in the Restore MCMC framework for light transport, outperforming prior methods and achieving real-time GPU performance.
-
Sim-FA: A GPGPU Simulator Framework for Fine-Grained Asynchronous Pipeline Analysis
A trace-driven simulator models Hopper TMA/WGMMA pipelines at WarpGroup granularity, matching H800 FlashAttention-3 latency to 5.7% MAPE, and its analytical model explains GenZ's long-sequence DRAM underestimation.
-
Reinforced Graph of Thoughts: RL-Driven Adaptive Prompting for LLMs
RGoT uses RL to adaptively generate task-specific graphs of operations for GoT-style LLM prompting from a human-provided set, with results suggesting feasibility under constraints.