2026

• posts Apr 22 Disaggregated Prefill: Splitting Compute Across Machines
• posts Apr 22 Prefix Caching: Reusing KV Cache Across Requests
• posts Apr 22 Chunked Prefill: Slicing the Prefill to Protect Decode Latency
• posts Apr 22 Continuous Batching: Scheduling at Iteration Granularity
• posts Apr 22 Paged Attention: Virtual Memory for the GPU
• posts Apr 21 Online Softmax: Tiling for Arbitrarily Large Rows
• posts Apr 20 Why KV Cache Works in LLM Inference
• posts Apr 20 Fused Softmax in Triton
• posts Apr 19 SSH Port Forwarding: Local and Remote Tunnels Explained
• posts Mar 22 Mitmproxy + Tampermonkey = better {llm, …} viewer
• posts Feb 16 Batch vs Stochastic Gradient Descent
• posts Feb 16 Forward & Backward Propagation

2024

• posts Jun 1 Key Management With GnuPG
• posts Feb 16 DSU on Tree (Sack)
• posts Feb 10 Shortest Paths Algorithms

2023

• posts Dec 19 Cycle Finding Algorithms
• posts Jan 21 Connect to your android wirelessly
• posts Jan 18 Tiling WM (i3)

2022

• posts Oct 21 Public, Private and Hybrid Cloud
• posts Apr 18 Use Random in C++

• 1
• 2
• 3