Topic
Hardware, silicon and the physical layer under the software
Software people are relearning that compute is a physical good with a supply chain, a power budget and a lead time. Chips, edge devices, sensing and the Shenzhen manufacturing base that makes most of it.
62 pieces
Teardowns
8- When a Knowledge Graph Beats a Vector Database Vector search retrieves what looks similar. It cannot tell you which shipped assemblies a bad plating bath touches two hops out. That is traversal, and traversal is a different operation.
- AI Visual Inspection Fails at the Lighting Long Before the Model Classical vision still wins geometric checks. Learned models earn their place on variable and rare defects. Accuracy is set by lighting, fixturing and the false-reject threshold you can afford.
- Six plants, one event log, and what it actually guarantees An event log across distributed plants buys ordering per partition, replayable history and one shared dependency, and most factories would be better served by a database for another five years.
- OT Security Breaks the IT Playbook, and AI Agents Make It Worse Availability outranks confidentiality, you cannot patch a machine mid-run, and Modbus has no password because in 1979 the wall was the password. Then we connected all of it to AI.
- Unified Namespace vs Data Historian: Why Neither Replaces the Other A UNS is a real-time topic space. A historian is an archive that discards data on purpose. The retention maths, the Sparkplug detail and the honest architecture where both exist.
- How DeepSeek Builds World-Class AI on a Shoestring Budget The real story behind the $5.6 million model that panicked Silicon Valley, explained so you actually understand it.
- Every insert writes a part Manufacturing telemetry is append-heavy, high-cardinality and almost always queried as a time-bounded aggregate. A columnar store fits that shape exactly, and breaks on updates, joins and small insert
- Generative AI for engineering design is cheap. Verification is where the money goes. The generative step in engineering costs minutes of compute. Proving the output manufacturable, simulation-clean and safe costs weeks, and that ratio decides which applications are worth doing.
Essays
8- Is AI a Bubble? Yes, and the Chips Will Still Be Running Hyperscaler capex reached roughly $745bn for 2026 while the two largest labs run at about $105bn combined. The bubble is real. The compute will outlive it. An engineer's read from Shenzhen.
- Made in China Quality: What the Label Actually Predicts in 2026 The label used to mean cheap and derivative. In batteries, solar, drones, robots and open-weight models it now marks the frontier. The perception lag is worth money to anyone testing on evidence.
- China Humanoid Robots Shipped 19,100 Units in Six Months. The Reducers Matter More. An ageing workforce and an explicit industrial policy point the same direction. Beneath the humanoid demos sits a component supply chain that quietly changed hands: reducers, sensors, actuators.
- Wheel-Legged Robots: Buying Two Locomotion Modes, Paying for Three Wheels are cheap per metre. Legs handle discontinuity. The hybrid tries to buy both and pays in actuators, distal mass and a switching problem. An engineer's account of when that trade is worth it.
- Shenzhen Hardware Manufacturing Is No Longer Cheap. It Is Still Four Days. Shenzhen stopped being cheap years ago. What it still sells is density: an iteration loop that runs in four days because the mould shop, the fab and the assembler are all inside an hour's drive.
- Seven China technology milestones that mattered and what each one cost Seven inflection points, from the 2008 Beijing-Tianjin line to DeepSeek R1, read for what each revealed about how decisions get made in China and who paid the bill. Triumph and dismissal both miss.
- Sensor Fusion in Sports Tracking and What 500Hz Is Actually For A 500Hz sensor in a football is a timestamping instrument, not a measuring one. What fusion actually buys you, and why clock alignment across devices breaks more systems than sampling rate ever does.
- AI in Supply Chain: The Dull Wins and the Forecasting Trap Forecasting gets the budget and moves the least. The gains that hold up in cross-border sourcing are duller: reading documents, watching lead times drift and shrinking the exception queue.
Signals
1Projects
2- How to Start a Tech Community in a City Where You Are Foreign The mechanics nobody writes down: why first meetings fail as coordination problems, why narrow definitions travel further than broad ones and what the Greater Bay Area gives an African engineer.
- Your sourcing agent is paid to make the order bigger Sourcing agents are paid a percentage of what you spend, so every good procurement decision costs them money. What engineer-led, fee-based buyer representation actually involves instead.
Tools
12 of 43- Gemini Notebook review Google's grounded research notebook, renamed from NotebookLM in July 2026. Upload sources, get answers with inline citations and run code against them in a per-notebook sandbox.
- Milvus review Apache 2.0 vector database from Zilliz, self-hostable from a pip install up to a billion-vector cluster. The open engine under Zilliz Cloud, minus the proprietary Cardinal search core.
- MuJoCo review Google DeepMind's rigid-body physics engine for robotics. Contact-rich dynamics, MJCF models, plus MJX and MuJoCo Warp for batched GPU rollouts during policy training.
- RAGFlow review InfiniFlow's Apache 2.0 RAG engine built on deep document parsing. DeepDoc handles layout, tables and OCR before chunking, and every citation points back to a highlighted region in the source PDF.
- RoboBrain 2.5 review BAAI's embodied brain model (悟界·RoboBrain), Apache 2.0 at 4B and 8B, shipped in parallel NVIDIA and Moore Threads builds. It plans and points in 3D for a controller underneath.
- Roboflow review End-to-end computer vision platform for labelling, training and deploying detection models to cloud or edge hardware like Jetson, with a visual builder for multi-step vision pipelines.
- Zilliz Cloud review Managed vector database built on Milvus, the open-source engine behind most billion-scale retrieval systems, for teams who need hybrid search without running the cluster themselves.
- pgvector review Postgres extension that adds vector types and approximate nearest neighbour indexes, so embeddings live in the same database as the rest of your data.
- Zhipu ChatGLM / 智谱清言 review hipu AI’s bilingual AI assistant powered by the GLM-4 model series, with strong academic roots from Tsinghua University and excellent bilingual capabilities
- Coze / 扣子 review ByteDance’s AI bot-building platform that enables anyone to create custom AI assistants with plugins, workflows, and knowledge bases without coding.
- Quark / 夸克 review Alibaba’s AI-powered search and productivity app combining intelligent search, document tools, cloud storage, and AI assistant capabilities.
- AgiBot World review Open manipulation data and policy weights from AgiBot (智元机器人): a million-plus real robot trajectories, the GO-1 VLA and the Genie Envisioner world model, all non-commercial.