My Homelab Journal

Unvieling the world of Deep Learning • Kernel Optimization • CUDA • AI Systems

HOME
CATEGORIES
TAGS
ARCHIVES
ABOUT

Home Tags bottleneck

Tag

bottleneck 1

My Journey Optimizing Attention: Why My First CUDA Optimization Barely Worked Aug 5, 2025

Recently Updated

From Confusion to Colors: My Journey Learning Parallel Image Magic with CUDA Threads
From Matrix Multiplication to Warp Optimizations — My Journey and Insights.
AI Action Item Extractor: Meeting Dialogue to JSON.
My Journey Optimizing Attention: Why My First CUDA Optimization Barely Worked
GPUs Are Lazy: Why Your Matrix Multiplication Is Wasting 37% Memory.

Trending Tags

gpu optimization performance profiling learning-journey matrix-multiplication parallel-computing shared-memory tiling warp-divergence

© 2025 Umang Singh. Some rights reserved.

Using the Chirpy theme for Jekyll.

Trending Tags

gpu optimization performance profiling learning-journey matrix-multiplication parallel-computing shared-memory tiling warp-divergence

A new version of content is available.