·20 min read
Continuous Dynamic Batching in LLM Inference: Orca, vLLM & TGI Latency Benchmarks
Benchmarking continuous dynamic batching algorithms: how Orca, vLLM, and TGI achieve 3.8x higher throughput on GPU inference clusters.
Technical Writings
In-depth systems architecture, distributed database benchmarks, and production engineering postmortems.
Top Topics
328Showing 41–50 of 310 articles
·22 min read
DeepSeek V3 Architecture Deep Dive: Multi-Head Latent Attention (MLA), DeepSeekMoE & FP8 Mixed Precision
Deep dive into DeepSeek V3 architecture: Multi-Head Latent Attention (MLA) KV compression, DeepSeekMoE routing, and FP8 mixed-precision training.
·25 min read
Distributed Locking in Distributed Systems: Redlock, PostgreSQL Advisory Locks & etcd Leases
Distributed locking tradeoffs compared: Redlock vs PostgreSQL advisory locks vs etcd raft leases under network partitions and split-brain scenarios.
·17 min read
High-Volume Distributed Tracing with OpenTelemetry, Kafka & ClickHouse
Ingesting 50,000 trace spans/sec with OpenTelemetry Collector, Kafka buffering, and ClickHouse columnar storage without losing spans or blowing budgets.
·15 min read
DuckDB-Wasm in Modern Web Apps: Client-Side OLAP, Parquet Streaming & Sub-Second Dashboards
Building sub-second interactive analytics dashboards with DuckDB-Wasm and Apache Arrow: streaming Parquet datasets entirely in the client browser.
·14 min read
eBPF & Cilium in Kubernetes: High-Throughput Routing, Network Policies & Hubble Observability
eBPF networking in Kubernetes with Cilium: bypassing iptables/kube-proxy, direct server return (DSR), and Hubble service-mesh flow inspection.
·15 min read
Open Table Formats in 2026: Apache Iceberg v2 vs Delta Lake 3.0 on Cloud Object Storage
Apache Iceberg vs Delta Lake benchmarked: ACID transaction isolation, schema evolution, hidden partitioning, and engine-agnostic query speeds on S3.
·16 min read
High-Performance Browser Storage: SQLite Wasm, Origin Private File System (OPFS) & Web Workers
Client-side storage architecture: comparing IndexedDB, OPFS (Origin Private File System), and SQLite WASM for local-first web applications.
·14 min read
Kubernetes Gateway API in Production: Migrating from Ingress-Nginx with Envoy
Migrating from ingress-nginx to Kubernetes Gateway API with Envoy: role-oriented routing, cross-namespace HTTPRoutes, and zero-downtime cutover.
·16 min read
Hierarchical Multi-Agent Systems in LangGraph: Supervisors, Subgraphs & State Machines
Multi-agent orchestration with LangGraph: hierarchical supervisor routing, shared state schemas, and subagent delegation patterns.
Showing 41-50 of 310 posts