Modular: How to Beat Unsloth's CUDA Kernel Using Mojo—With Zero GPU Experience
From the article
You might also wanna read
Unsloth and NVIDIA Partner to Accelerate LLM Fine-Tuning by 20%
Unsloth has partnered with NVIDIA to optimize fine-tuning of large language models, achieving 20% faster training speeds. The collaboration
Unsloth - Train and Run Models Locally·2mo ago
GPU-Optimized Datalog Evaluation: GPULOG System Analysis from ASPLOS'25 Paper
This article analyzes the ASPLOS'25 paper 'Optimizing Datalog for the GPU,' which presents GPULOG, a system that optimizes Datalog evaluatio
danglingpointers.substack.com·8mo ago
APEX4: Platform-Dependent W4A4 LLM Inference via Intra-SM Compute Rebalancing
This paper presents APEX4, a system for efficient W4A4 (4-bit weights, 4-bit activations) LLM inference that addresses the bottleneck of gro
arxiv.org·29d ago
AI-Driven Approach for Portable GPU Kernels in High-Performance Computing
This academic paper from North Carolina State University researchers presents an approach to leveraging AI ecosystems for creating portable
hgpu.org·22d ago
cuTile Rust: Extending Rust's Ownership Model to Safe GPU Kernel Programming
This article presents cuTile Rust, a system that extends Rust's ownership and memory safety guarantees to GPU kernel programming. It allows
arxiv.org·21d ago
Performance Analysis of Open-Source NVIDIA Linux Drivers: Nouveau+NVK for Gaming and Compute
This article provides an in-depth performance analysis of the open-source NVIDIA Linux driver stack, specifically examining the Nouveau kern
phoronix.com·7mo ago
Comments
Sign in to join the conversation.
No comments yet. Be the first.