·
DataBubble
  • Home
  • Models
  • News
  • Compare
  • Boards
  • Pricing
  • About
  • Newsletter
  • Methodology
  • Contact
Latest
Benford's Law as a Distributional Prior for Post-Training Quantization of Large Language Models6h◆Transformers as In-Context Samplers: From Closed-Form Diffusion to Estimation-Free Sampling6h◆Evidence-Aligned Local Composition of Discrete Experts for Sequence Restoration6h◆Dynin-Robotics: Omnimodal Unified Diffusion Vision-Language-Action Model6h◆FEAT: A Linear-Complexity Foundation Model for Extremely Large Structured Data6h◆A Multi-Vehicle Dataset with Camera, LiDAR, and Radar Sensors and Scanned 3D Models for Custom Auto-Annotation using RTK-GNSS6h◆Autonomous Research for Open-Ended Problems: A Case Study on Telecom Ticket Retrieval6h◆Physics-enriched neural solvers for transient ice-flow simulation6h◆Dissecting GPU Utilization for LLM Inference on Nvidia Hopper6h◆Scaling Online Complex Event Detection with Synthetic Supervision and Mamba-Based Neural Algorithmic Reasoning6h◆Space as an Interventional Invariant: Cross-Modal Predictive Geometry for Stratified Cities and Em-Spaced Intelligence6h◆Cortex: Content Analysis Support Software, a Resource for Qualitative Research6h◆Is Bash All You Need? An Empirical Study of Tool Interfaces for Enterprise Digital Worker Agents6h◆KoSimpleQA: A Korean Factuality Benchmark with an Analysis of Reasoning LLMs6h◆Translationese as a Rational Response to Translation Task Difficulty6h◆Can Foundation Models Moderate Online Content? Evaluating Instruction- vs. Example-Driven Policy Operationalization6h◆Inverting Self-Triggered Control: Adversarial Reinforcement Learning for Sparse Denial-of-Service Attacks6h◆Classical and quantum kernel fusion for two-sample testing6h◆Granularity-Adaptive Credit Assignment for Long-Horizon LLM Agent Reinforcement Learning6h◆One Simple Trick for Improving the Performance of Energy-Limited Local Inference and Training6h◆Benford's Law as a Distributional Prior for Post-Training Quantization of Large Language Models6h◆Transformers as In-Context Samplers: From Closed-Form Diffusion to Estimation-Free Sampling6h◆Evidence-Aligned Local Composition of Discrete Experts for Sequence Restoration6h◆Dynin-Robotics: Omnimodal Unified Diffusion Vision-Language-Action Model6h◆FEAT: A Linear-Complexity Foundation Model for Extremely Large Structured Data6h◆A Multi-Vehicle Dataset with Camera, LiDAR, and Radar Sensors and Scanned 3D Models for Custom Auto-Annotation using RTK-GNSS6h◆Autonomous Research for Open-Ended Problems: A Case Study on Telecom Ticket Retrieval6h◆Physics-enriched neural solvers for transient ice-flow simulation6h◆Dissecting GPU Utilization for LLM Inference on Nvidia Hopper6h◆Scaling Online Complex Event Detection with Synthetic Supervision and Mamba-Based Neural Algorithmic Reasoning6h◆Space as an Interventional Invariant: Cross-Modal Predictive Geometry for Stratified Cities and Em-Spaced Intelligence6h◆Cortex: Content Analysis Support Software, a Resource for Qualitative Research6h◆Is Bash All You Need? An Empirical Study of Tool Interfaces for Enterprise Digital Worker Agents6h◆KoSimpleQA: A Korean Factuality Benchmark with an Analysis of Reasoning LLMs6h◆Translationese as a Rational Response to Translation Task Difficulty6h◆Can Foundation Models Moderate Online Content? Evaluating Instruction- vs. Example-Driven Policy Operationalization6h◆Inverting Self-Triggered Control: Adversarial Reinforcement Learning for Sparse Denial-of-Service Attacks6h◆Classical and quantum kernel fusion for two-sample testing6h◆Granularity-Adaptive Credit Assignment for Long-Horizon LLM Agent Reinforcement Learning6h◆One Simple Trick for Improving the Performance of Energy-Limited Local Inference and Training6h◆
News/Dissecting GPU Utilization for LLM Inference on Nvidia Hopper
arxiv
PublishedSeptember 14, 2026 at 4:00 AM
—neutral

Dissecting GPU Utilization for LLM Inference on Nvidia Hopper

Source
arxiv.orgfull article ↗
Read on arxiv→
Publisher summary· verbatim

arXiv:2609.12923v1 Announce Type: cross Abstract: A single SM utilization percentage can make an LLM inference workload look compute-saturated while hiding how much useful work is being done. The problem is not that the counter is wrong, but that it collapses several different mechanisms into one nu

Stay posted· Newsletter

A 5-min weekly brief — top movers, price watch, story of the week.

// no spam · unsubscribe one-click · free forever

Discussion
Source
↗
arxiv
Read original ↗All from arxiv →

No replies yet. Be first.

Source
↗
arxiv
Read original ↗All from arxiv →

Related coverage

More from ARXIV
arxivBenford's Law as a Distributional Prior for Post-Training Quantization of Large Language Models6harxivTransformers as In-Context Samplers: From Closed-Form Diffusion to Estimation-Free Sampling6harxivEvidence-Aligned Local Composition of Discrete Experts for Sequence Restoration6harxivDynin-Robotics: Omnimodal Unified Diffusion Vision-Language-Action Model6h
The Bubble Brief
WEEKLY

Read AI insights every Tuesday — top movers, new releases, story of the week.

// no spam · unsubscribe one-click · free forever

Originally published on arxiv ↗
HomeModelsNews