Triton
PulseAugur coverage of Triton — every cluster mentioning Triton across labs, papers, and developer communities, ranked by signal.
6 day(s) with sentiment data
-
New attention quantization speeds up tabular foundation models
Researchers have developed a new attention quantization strategy for tabular foundation models to improve inference performance. This method focuses on quantizing queries, keys, and values to FP8, leveraging explicit FP…
-
New tools and research tackle GPU optimization for AI workloads
Several research papers and a new open-source tool address challenges in optimizing AI workloads on GPUs. COMPASS-ABS aims to reduce fragmentation in shared GPU clusters for deep learning training, improving resource ut…
-
AI and HPC converge via Stencil Computations with Triton and SYCL
This article discusses the intersection of AI and high-performance computing (HPC) through the lens of stencil computations. It highlights how techniques like vector outer products and the Triton programming model can b…
-
Einsummable package enables automatic multi-GPU parallelism for AI
A new package called Einsummable has been developed to automatically parallelize AI computations across multiple GPUs. This approach leverages CUDA and Triton to enable efficient multi-GPU execution for large language m…
-
New AI agent AdaExplore boosts kernel code generation efficiency
Researchers have developed AdaExplore, a novel agent framework designed to improve the generation of efficient kernel code, particularly for domain-specific languages like Triton. The system employs a two-stage process:…
-
BEAM3R uses Mamba-3 for faster radiation dose reconstruction
Researchers have developed BEAM3R, a novel framework for accurate and rapid dose calculation in radiation therapy. This system utilizes the Mamba-3 state-space model, eschewing expensive 3D convolutions for a more effic…
-
New ebook guides ML practitioners on optimizing GPU performance with Triton
Manning Publications has released an early-access ebook, "GPU Programming with Triton," by Harshwardhan Fartale. The book aims to help machine learning practitioners optimize their models by writing custom GPU kernels i…
-
LoGo introduces dynamic attention for efficient long-context LLMs
Researchers have introduced LoGo, a novel token-level dynamic local-global attention mechanism designed to improve the efficiency of large language models. LoGo dynamically allocates attention budgets by allowing all to…
-
New benchmark tests LLMs for GPU database query optimization
Researchers have developed DataKernelBench, a new benchmark designed to evaluate the ability of large language models (LLMs) to optimize database queries specifically for GPUs. Unlike existing benchmarks that focus on m…
-
New rl-triton library speeds up RL credit assignment with unified GPU kernels
Researchers have developed rl-triton, an open-source library featuring high-performance GPU kernels for reinforcement learning credit assignment. This library unifies seven distinct RL estimation algorithms, including G…
-
Understanding GPU Kernels: The Key to Efficient AI Inference
The article explains the concept of a 'kernel' in the context of GPU computation for AI models, defining it as a single function executed by thousands of threads in parallel. It highlights that the primary challenge in …
-
New frameworks optimize GPU kernels for deep learning and HPC · 3 sources tracked
Three new research papers introduce advanced frameworks for optimizing GPU kernels, crucial for deep learning and high-performance computing. HIERA focuses on workload-aware planning across different implementation spac…
-
Huawei open-sources AscendNPU IR, enhancing BiSheng compiler flexibility
Huawei has open-sourced the AscendNPU IR, a core component of its BiSheng compiler. This MLIR-based abstraction layer is designed for Ascend hardware, offering support for front-end languages like Triton and extending c…
-
New Benchmark Tests LLMs on Real-World Triton Kernel Generation
Researchers have introduced RealisticTritonBench, a new benchmark designed to evaluate the ability of large language models (LLMs) to generate Triton kernels for AI frameworks. Unlike previous benchmarks, RealisticTrito…
-
OpenAI cuts GPT-5.6 prices by up to 80% with AI-driven optimizations
OpenAI has significantly reduced the prices for its GPT-5.6 model family, with the Luna tier now 80% cheaper and the Terra tier 20% less expensive. This cost reduction was largely driven by engineering optimizations, in…
-
Rust and CUDA C++ outperform Triton on irregular GPU workloads
A new research paper compares the performance of CUDA C++, Rust, and Triton for GPU workloads, particularly focusing on irregular operations like hash table insertions. The study found that while all three languages per…
-
Attn-QAT enables stable 4-bit attention training for LLMs
Researchers have developed Attn-QAT, a novel method for 4-bit quantization-aware training of attention mechanisms in large language models. This approach addresses the challenges of low precision in FP4 computation, par…
-
New linearized attention model achieves higher accuracy and lower perplexity
Researchers have developed a linearized version of 2-simplicial attention, which rewrites the trilinear score into an inner product. This new form allows for linear cost in sequence length while maintaining global reach…
-
New method retrofits linear attention to speed up diffusion language models
Researchers have developed a method to retrofit linear attention into diffusion language models (dLLMs) to accelerate inference. This new approach, called block-hybrid attention, combines exact softmax attention within …
-
Open-source Triton driver enables DirectX 11 on QEMU VMs
A new open-source driver called Triton has been developed that enables DirectX 11 to run within QEMU virtual machines. This advancement is part of a broader trend towards integrating AI into the development process, wit…