REVIEW 32 cited by
TileLang: A Composable Tiled Programming Model for AI Systems
Not yet reviewed by Pith; the record is open.
This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.
SPECIMEN: schema-true, not a live event
T0 review · schema-true
One-sentence machine reading of the paper's core claim.
pith:XXXXXXXX · record.json · timestamp
TileLang: A Composable Tiled Programming Model for AI Systems
read the original abstract
Modern AI workloads rely heavily on optimized computing kernels for both training and inference. These AI kernels follow well-defined data-flow patterns, such as moving tiles between DRAM and SRAM and performing a sequence of computations on those tiles. However, writing high-performance kernels remains complex despite the clarity of these patterns. Achieving peak performance requires careful, hardware-centric optimizations to fully leverage modern accelerators. While domain-specific compilers attempt to reduce the burden of writing high-performance kernels, they often struggle with usability and expressiveness gaps. In this paper, we present TileLang, a generalized tiled programming model for more efficient AI Kernel programming. TileLang decouples scheduling space (thread binding, layout, tensorize and pipeline) from dataflow, and encapsulated them as a set of customization annotations and primitives. This approach allows users to focus on the kernel's data-flow itself, while leaving most other optimizations to compilers. We conduct comprehensive experiments on commonly-used devices, across numerous experiments, our evaluation shows that TileLang can achieve state-of-the-art performance in key kernels, demonstrating that its unified block-and-thread paradigm and transparent scheduling capabilities deliver both the power and flexibility demanded by modern AI system development.
Forward citations
Cited by 32 Pith papers
-
Step-TP: A Grounded, Step-Level Dataset with Chain-of-Thought Reasoning for LLM-Guided Tensor Program Optimization
Step-TP is a dataset providing grounded, atomic step-level IR transitions and CoT supervision to enable reliable multi-step LLM-guided tensor program optimization instead of end-to-end imitation.
-
Characterizing Real-World Bugs in Tile Programs for Automated Bug Detection
A taxonomy of 301 real tile-program codegen bugs: 48.8% are type/operator bugs, 19.3% are memory bugs, and detection needs shape-aware fuzzing plus differential oracles.
-
Characterizing Real-World Bugs in Tile Programs for Automated Bug Detection
A systematic analysis of 301 real-world code generation bugs in tile programs, categorizing root causes, symptoms, input patterns, test oracles, and fix strategies from curated GitHub reports.
-
CODA: Rewriting Transformer Blocks as GEMM-Epilogue Programs
CODA re-expresses non-attention Transformer computations as GEMM epilogues using composable primitives for scaling, reductions, and transformations.
-
CODA: Rewriting Transformer Blocks as GEMM-Epilogue Programs
CODA re-expresses most non-attention Transformer computations as GEMM-plus-epilogue programs using a constrained set of composable primitives to keep intermediate results on-chip and cut global memory traffic.
-
Efficient Video Diffusion Models: Advancements and Challenges
A survey that groups efficient video diffusion methods into four paradigms—step distillation, efficient attention, model compression, and cache/trajectory optimization—and outlines open challenges for practical use.
-
Nautilus: An Auto-Scheduling Tensor Compiler for Efficient Tiled GPU Kernels
Nautilus auto-compiles math-like tensor descriptions into optimized GPU kernels, delivering up to 42% higher throughput than prior compilers on transformer models across NVIDIA GPUs.
-
A Full-Stack Performance Evaluation Infrastructure for 3D-DRAM-based LLM Accelerators
ATLAS is the first silicon-validated simulation framework for 3D-DRAM LLM accelerators, achieving under 8.57% error and over 97% correlation with real hardware while supporting design exploration.
-
GPU-Tile-Sim: A Tile-Centric GPU Simulation Framework for LLM Hardware-Software Co-Design
A tile-graph GPU simulator predicts optimized LLM kernel performance on A100/H100 with MAPE about 1–9% by modeling dependencies and overlap rather than full instruction pipelines.
-
RotateAttention: RoPE-Aware Rotation and Range Rectification for INT4 Quantized Attention in Video Generation
RoPE-aware orthogonal rotations plus full-range INT4 quantization of attention probabilities deliver mixed-precision FlashAttention that matches FP16 video quality at 1.5–1.68× end-to-end speedup.
-
RotateAttention: RoPE-Aware Rotation and Range Rectification for INT4 Quantized Attention in Video Generation
RoPE-aware 2×2 rotations plus fixed-scale affine INT4 quantization of P enable mixed-precision FlashAttention for 3D-RoPE video DiTs with near-FP16 quality and 1.5–1.7× end-to-end speedup.
-
Axon: A Synthesizing Superoptimizer for Tensor Programs
Axon automatically synthesizes high-performance tensor programs for AI accelerators via SMT-guaranteed semantic transformations, hardware-constrained tiling, and empirical performance selection.
-
MiniMax Sparse Attention
MiniMax Sparse Attention is a GQA-based block-sparse attention mechanism that selects top-k blocks independently per group and delivers 28.4x per-token compute reduction at 1M context with on-par performance plus 14.2...
-
Veda: Scalable Video Diffusion via Distilled Sparse Attention
Veda formulates tile selection in video diffusion attention as a reconstruction problem from full attention maps, using statistics-aware and head-aware scoring to enable high sparsity with maintained quality and hardw...
-
TLX: Hardware-Native, Evolvable MIMW GPU Compiler for Large-scale Production Environments
TLX is a Triton extension that exposes multi-warp, asynchronous, and cluster-level controls for modern GPUs, delivering competitive performance with low programmer effort and production deployment.
-
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-...
-
FalconGEMM: Surpassing Hardware Peaks with Lower-Complexity Matrix Multiplication
FalconGEMM delivers a framework with deployment, group-parallel execution, and analytical decision modules that makes lower-complexity matrix multiplication practical, beating cuBLAS and similar libraries by 7.59-17.8...
-
FalconGEMM: Surpassing Hardware Peaks with Lower-Complexity Matrix Multiplication
FalconGEMM is a framework with deployment, execution, and decision modules that makes lower-complexity matrix multiplication practical, outperforming standard GEMM libraries by 7.59-17.85% and competitors like AlphaTe...
-
LIVEditor-14B: Lightning Unified Video Editing via In-Context Sparse Attention
LIVEditor-14B applies a new sparse attention method (ISA) that prunes context and uses query-sharpness routing to cut attention latency ~60% with no loss in editing quality on standard benchmarks.
-
DeepStack: Scalable and Accurate Design Space Exploration for Distributed 3D-Stacked AI Accelerators
DeepStack introduces a fast performance model and hierarchical search method for co-optimizing 3D DRAM stacking, interconnects, and distributed scheduling in AI accelerators, delivering up to 9.5x throughput gains ove...
-
HISA: Efficient Hierarchical Indexing for Fine-Grained Sparse Attention
HISA speeds up fine-grained sparse attention indexers via block-then-token hierarchy, delivering substantial speedups at 64K context with no training and quality matching the original DSA on long-context benchmarks.
-
Three non-Hermitian random matrix universality classes of complex edge statistics: Spacing ratios and distributions
Three non-Hermitian edge universality classes show distinct repulsion, universal cubic small-spacing behaviour, and incomplete unfolding by complex spacing ratios at the edge.
-
mHC: Manifold-Constrained Hyper-Connections
mHC projects hyper-connection residual spaces onto a manifold to restore identity mapping, enabling stable large-scale training with performance gains over standard HC.
-
TileLoom: Automatic Dataflow Planning for Tile-Based Languages on Spatial Dataflow Accelerators
TileLoom compiles tile-based languages to spatial dataflow hardware by distributing tiles across cores and optimizing data reuse via on-chip networks, delivering vendor-comparable performance on Tenstorrent systems.
-
HTAM: Hierarchical Transition-Attended Memory for Operator Optimization
HTAM builds a Hierarchical Transition Graph to organize coarse global directions and detailed local strategies for guiding LLM-based CUDA kernel optimization, improving results on KernelBench.
-
LIVEditor-14B: Lightning Unified Video Editing via In-Context Sparse Attention
ISA prunes low-saliency context tokens and routes queries by sharpness to either full or 0-th order Taylor sparse attention, enabling LIVEditor to cut attention latency ~60% while beating prior video editing methods o...
-
StreamIndex: Memory-Bounded Compressed Sparse Attention via Streaming Top-k
Chunked streaming top-k enables CSA indexer execution at 1M sequence length with 6.21 GB peak memory and >=0.998 recall on synthetic V4-shaped inputs.
-
HieraSparse: Hierarchical Semi-Structured Sparse KV Attention
HieraSparse delivers a hierarchical semi-structured sparse KV attention system that achieves 1.2x KV compression and 4.57x decode attention speedup versus prior unstructured sparsity methods at equivalent sparsity, pl...
-
Understand and Accelerate Memory Processing Pipeline for Large Language Model Inference
Unifying LLM memory optimizations into a Prepare-Compute-Retrieve-Apply pipeline and accelerating it on GPU-FPGA hardware yields up to 2.2x faster inference and 4.7x less energy than GPU-only baselines.
-
Understand and Accelerate Memory Processing Pipeline for Large Language Model Inference
Unifying LLM long-context optimizations into a four-step memory pipeline and offloading sparse/irregular stages to FPGA yields up to 2.2× speedup and 4.7× energy savings versus GPU-only.
-
AscendOptimizer: Episodic Agent for Ascend NPU Operator Optimization
AscendOptimizer combines kernel rewinding for reusable experience with evolutionary search on hardware feedback to optimize Ascend NPU operators, delivering 1.21x geometric-mean speedup and faster performance on 53.47...
-
DMuon: Efficient Distributed Muon Training with Near-Adam Overhead
DMuon delivers 1.48x-3.01x end-to-end and 6.85x-163x optimizer-step speedups for Muon on embodied foundation models and LLMs while matching AdamW per-step latency.
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.