Topics & Research Index
Discover peer-reviewed architectural analyses and performance benchmarks organized by specific systems, runtimes, and engineering disciplines.
AI Systems
Latest: Inside OpenAI Dots: Architecture of Persistent Autonomous Cloud Agents
Distributed Systems
Latest: Deterministic Consensus at Global Scale: Engineering Multi-Region Quorum Topologies
Autonomous Agents
Latest: Inside OpenAI Dots: Architecture of Persistent Autonomous Cloud Agents
Cloud Architecture
Latest: Inside OpenAI Dots: Architecture of Persistent Autonomous Cloud Agents
Cloud Infrastructure
Latest: Deterministic Consensus at Global Scale: Engineering Multi-Region Quorum Topologies
Consensus
Latest: Deterministic Consensus at Global Scale: Engineering Multi-Region Quorum Topologies
CUDA
Latest: Sub-Millisecond Token Scheduling: Inside Disaggregated KV-Cache Paging for Frontier LLMs
Enterprise AI
Latest: Inside OpenAI Dots: Architecture of Persistent Autonomous Cloud Agents
GPT-6 Astra
Latest: Inside OpenAI Dots: Architecture of Persistent Autonomous Cloud Agents
GPU Kernels
Latest: Sub-Millisecond Token Scheduling: Inside Disaggregated KV-Cache Paging for Frontier LLMs
High Availability
Latest: Deterministic Consensus at Global Scale: Engineering Multi-Region Quorum Topologies
Linux Kernel
Latest: Zero-Copy Kernel Datapaths: Building Thread-Per-Core Rust Services with io_uring
LLM Inference
Latest: Sub-Millisecond Token Scheduling: Inside Disaggregated KV-Cache Paging for Frontier LLMs
Multi-Region
Latest: Deterministic Consensus at Global Scale: Engineering Multi-Region Quorum Topologies
Networking
Latest: Zero-Copy Kernel Datapaths: Building Thread-Per-Core Rust Services with io_uring
OpenAI
Latest: Inside OpenAI Dots: Architecture of Persistent Autonomous Cloud Agents
Performance
Latest: Zero-Copy Kernel Datapaths: Building Thread-Per-Core Rust Services with io_uring
Rust
Latest: Zero-Copy Kernel Datapaths: Building Thread-Per-Core Rust Services with io_uring
Systems Programming
Latest: Zero-Copy Kernel Datapaths: Building Thread-Per-Core Rust Services with io_uring