Hardware Tools

Motorola Razr Fold is Motorola’s first book-style foldable, with an 8.1-inch inner display, Snapdragon 8 Gen 5, 6000mAh battery, 80W charging, a DXOMARK Gold-rated foldable camera system, and Moto Pen Ultra support for notes, annotations, and big-screen productivity.

Hardware is still built with processes that are 10-100 years old. Software got layers of abstraction decades ago while hardware never did. Atech is Lego for real electronics. Snap modules together, describe what you want it to do, and we generate the firmware. Idea to working device in minutes. No datasheet deep-dives, no soldering, no wondering "why doesn't it work?"

Framework Laptop 13 Pro pairs Intel Core Ultra Series 3 processors with a 21% larger battery for real 20-hour life, LPCAMM2 memory you can upgrade, and a refined CNC aluminum chassis. It keeps the repairable and modular DNA that made Framework famous.

Typing is the bottleneck. SpeakON removes it. A MagSafe AI device for iPhone — press once to speak into any app. No mic permission. No switching. Even works with your phone locked. Zero friction.

Embedist is a Windows desktop application that combines AI assistance with embedded firmware development. Built with Tauri 2, React, and TypeScript, it brings board-aware AI debugging, real-time serial monitoring, and PlatformIO build integration into a single cohesive environment.

Claude Desktop Buddy exposes a lightweight BLE API from the Claude desktop app. It allows makers to connect Claude Cowork and Claude Code to physical microcontrollers (like the ESP32) to display states, handle permission prompts, and more.

Zuflow is a web app that helps you build 3D assemblies with visual logic. Map your parts on a node graph and let the engine do the math. 🕸️ Visual Constraints: Wire parts on a canvas. See your logic. ⚡ Auto-Solver: Define relationships and watch parts snap together. 🚀 Browser-Native: Fast 3D rendering with zero downloads. 📐 Instant Drawings: Export dimensioned PDFs and DXFs

DJI Osmo Pocket 4 pushes the pocket gimbal camera into a much more serious imaging system: 1-inch CMOS, 4K/240fps, 14-stop dynamic range, 10-bit D-Log, 2× lossless zoom, built-in storage, and faster pro-grade workflow features in a device that still fits in your pocket.

Refaire helps hardware companies automate product technical support when issues require physical diagnosis. Customers' "facetime" to show and describe the problem, and Refaire uses vision, voice, and product knowledge to diagnose issues, guide troubleshooting, recommend the right parts or services, and escalate to a human with full session context.

The Arm AGI CPU is production silicon for AI infrastructure, delivering high performance and extreme density for agentic AI in modern data centers.

A set of advanced theoretically grounded quantization algorithms that enable massive compression for large language models and vector search engines.

Everest C1 is the world’s most powerful ARM64 server APU, built for engineers who demand uncompromising performance. Driven by the Snapdragon X2 Elite Extreme with 18 cores at 5 GHz, 128 GB unified memory, 80+ AI TOPS NPU, and enterprise IPMI. Silent, under 75 W, and Kubernetes‑native. The C1 delivers AI inference, edge clusters, and sovereign infrastructure at a fraction of cloud costs.

Meta is accelerating its custom silicon roadmap with four new MTIA chips in two years. Built with an inference-first focus and native PyTorch integration, they are designed to cost-effectively power GenAI at a massive consumer scale.

Manufacturing engineers spend hours pushing changes through PLM systems. Not engineering. Just hunting part numbers, checking BOMs, pulling cost data, and formatting ECRs. Sutra runs engineering change on autopilot. Describe the problem and Sutra makes the change decision, verifies parts, analyzes BOM impact, generates the change packet, and routes approvals automatically. What took hours now takes minutes.

VENTUNO Q is the first single-board computer born from Qualcomm's acquisition of Arduino. Fusing a 40 TOPS Snapdragon NPU with an STM32 microcontroller creates a unified dual-brain architecture, perfectly balancing heavy AI inference with real-time robotics.

The Nothing Phone (4a) Pro features a slim 7.95 mm full-metal unibody, confining its signature transparency to the camera module. Running the Snapdragon 7 Gen 4, the phone delivers a 3000-nit Glyph Matrix and up to 140x telephoto zoom.

Introducing MacBook Neo — Apple's all-new MacBook features a durable aluminum design, a stunning 13-inch Liquid Retina display, the power of Apple silicon, and all-day battery life — all for the breakthrough starting price of just $599

Introducing the new MacBook Neo. Stunning colours, durable aluminium design, Liquid Retina display, all-day battery life, AI and Apple Intelligence.

Hi Product Hunt 👋 I’m the founder of PI SpeedLAB. We started building PI-Link because we were frustrated with traditional speed guns — limited processing power, inconsistent readings, and expensive hardware. So we asked: What if we used the smartphone’s CPU instead? By combining Doppler radar hardware with smartphone-level computing power, we built a smarter way to measure baseball velocity. We’d love your feedback — especially from athletes, coaches, and hardware builders. Ask me anything!

Home cloud server with 4 NVMe SSD slots for up to 16TB storage. Milled from a single block of aluminum and framed with American Walnut. Powered by umbrelOS - run OpenClaw, Immich (photo/video backups), and hundreds of self-hosted apps with one click.

Forge turns PyTorch models into optimized CUDA and Triton kernels automatically. 32 AI agents run in parallel, each trying different optimization strategies like tensor cores, memory coalescing, and kernel fusion. A judge validates every kernel for correctness before benchmarking. We got 5x faster inference than torch.compile on Llama 3.1 8B and 4x on Qwen 2.5 7B. Works on any PyTorch model. Free trial on one kernel. Full credit refund if we don't beat torch.compile.