KernelOPT: Dispatch-Aware Agentic Search for GPU Kernel Optimization

Xtrace: High-Fidelity GPU Intra-Kernel Tracing via Binary-Level Instruction Splicing

Microarchitectural Memory Bandwidth Saturation, KV-Cache Paging Dynamics, and Time-to-First-Token Latency: A Comparative Benchmark of vLLM, TensorRT-LLM, and FlashAttention-3 on NVIDIA Hopper H100 ver...

Accelerating the Mitigation of LLM Inference Nondeterminism Across GPU Architectures

pytest-gpu-proof: Enabling Cloud-CPU Continuous Integration for GPU Code with Local GPU Attestation

DeepSeek-V4-Flash on AMD gfx90a: Correctness Recovery and Inference Performance Engineering

Automated Instruction Encoding Synthesis for Modern GPU ISA Compression

PrefixBench-H100: Characterizing Prefix Reuse and Time-to-First-Token in H100 LLM Serving

Accelerating the Solving of Many Tiny General Linear Systems on GPUs: Application to Constitutive Laws

AutoTuneBench: Trustworthy Measurement for Agent Auto-Tuning of LLM Serving Engines

Stencil Computation at the Intersection of AI and HPC

MaxKernel: Agentic Kernel Generation for TPUs

Taming Bitwise Behavior in GPU Kernels with Tensor Core: Black-Box Reconstruction, Compiler Enforcement, and Static Verification

Every Kernel Is a Join: Automatic Multi-GPU Parallelism for AI Computations in Einsummable

Hardware-Aware FP4 FlashAttention-4

PTXBench: Benchmark and Adapt LLMs for GPU Kernel Optimization with Architecture-specific PTX

Benchmarking Confidential Computing Performance on NVIDIA Blackwell GPUs

Concurrency Response of Plain Global Loads on the NVIDIA H100

An HPC Approach to Accelerate Tensor Decompositions

RealisticTritonBench: A Benchmark for Triton-Kernel Generation in Real-World AI Frameworks

Hand-Written PTX Tensor-Core GEMM Kernels: A Multi-Precision Study on NVIDIA L4

A Contract-Grade Verifier for LLM-Generated GPU Kernels, and a Native Blackwell Backward for the Gated-Linear-Recurrence Family

Spec Sheets Are Not Kernels: An ISA- and Source-Level Audit of INT8 Availability on NVIDIA Blackwell Ultra

CAKE: Compiler-Agent Co-Design for Frontier Kernel Evolution

Validation-Centric AI-Assisted GPU Porting of a 250,000+ Line Legacy Weather Simulation Code

Compiler-Grounded Hierarchical Diagnosis for LLM-Based Triton Kernel Optimization

Harness Engineering for LLM-Driven GPU Kernel Generation

PortLBM: A Portable Lattice Boltzmann Tool Leveraging SYCL on AMD, NVIDIA, and Intel GPUs

FlashPDE: A Drop-In Fused Triton Operator Library for Neural PDE Solvers

NVIDIA-labs OO Agents: Native Python Object-Oriented Agents

Enhancing the Performance Analysis of NCCL GPU Collectives

Real FP4 Tensor-Core Code in Pure Rust on a Gaming GPU – with NVIDIA’s Own Compiler

UniCoder: Unified Visual-to-Code Generation via Symbolic Rewards and Reference-Guided Code Optimization

Augmenting LLM Code Translation with Compiler Analysis for C to Triton Kernel Generation

CuFuzz: An API-Knowledge-Graph Coverage-Driven Fuzzing Framework for CUDA Libraries

Optimizing CUDA like a Human: Micro-Profiling Tools as Expert Surrogates for LLM-Based GPU Kernel Optimization

SpecGen: Accelerating Agentic Kernel Optimization with Speculative Generation

The Correctness Illusion in LLM-Generated GPU Kernels

Probe-and-Refine Tuning of Repository Guidance for Coding Agents

AutoPass: Evidence-Guided LLM Agents for Compiler Performance Tuning

daVinci-kernel: Co-Evolving Skill Selection, Summarization, and Utilization via RL for GPU Kernel Optimization

Fearless Concurrency on the GPU

Tangram: Hiding GPU Heterogeneity for Efficient LLM Parallelization

From Tokens to Regions: CUDA-Sensitive Instruction Tuning for GPU Kernel Generation

Leveraging AI Ecosystem for Portable and Sustainable GPU Kernels in HPC

Towards Feedback-to-Plan Decisions for Self-Evolving LLM Agents in CUDA Kernel Generation

CodegenBench: Can LLMs Write Efficient Code Across Architectures?

KForge: LLM-Driven Cross-Platform Kernel Generation for AI Accelerators

MusaCoder: Native GPU Kernel Generation with Full-Stack Training on Moore Threads GPU

Autonomous heterogeneous catalyst discovery with a self-evolving multi-agent digital twin