IEEE

To Tile or not to Tile, That is the Question

Abstract: A wide variety of graph algorithms expressed as linear algebra operations, i.e., triangle counting, k-truss analysis, breath first search, betweenness centrality, depend on the masked sparse ...
Tech Xplore

AI models stumble on basic multiplication without special training methods, study finds

These days, large language models can handle increasingly complex tasks, writing complex code and engaging in sophisticated ...
IEEE

Multiplication-Free Lookup-Based CNN Accelerator Using Residual Vector Quantization and Its FPGA Implementation

Abstract: In this paper, a table lookup-based computing technique is proposed to perform convolutional neural network (CNN) inference without multiplication, and its FPGA implementation is ...
GitHub

CUDA-L2: Surpassing cuBLAS Performance for Matrix Multiplication through Reinforcement Learning

CUDA-L2 is a system that combines large language models (LLMs) and reinforcement learning (RL) to automatically optimize Half-precision General Matrix Multiply (HGEMM) CUDA kernels. CUDA-L2 ...