🍎 macOS Hardware Info

Professional Hardware Diagnostics, Multi-Drive Benchmarking & AI Silicon Performance Suite for macOS

High-precision hardware telemetry, physical drive I/O benchmarking, and Apple Silicon / Intel compute stress testing unified into a single standalone tool.

Engineered by samuraidev • Portfolio

Key Capabilities • Architecture • CLI & Interactive Usage • Storage Engine • Power Subsystem • AI Benchmarks

---

🎯 Key Capabilities

macOS-Hardware-info.py delivers unified, low-level diagnostics and scientific hardware telemetry across Intel (T2) and Apple Silicon (M1/M2/M3/M4) architectures:

1. Modular Diagnostic Architecture — Interactive console menu with 5 targeted workflows and dedicated CLI flags for headless automation.
2. Comprehensive Storage Diagnostics & Speed Testing — S.M.A.R.T. health, TBW tracking, and unbuffered parallel I/O speed measurement for internal SSDs and external drives (USB-C, Thunderbolt, UASP).
3. Dual-Layer Battery & Power Telemetry — Full parity with official macOS Battery Health Management (BHM) paired with raw IOKit electrochemical data (design mAh, nominal capacity, cycles, cell temperature, and voltage).
4. AI Silicon Stress Testing — Real hardware compute benchmarks across CPU (FP32 GFLOPS via BLAS), GPU (FP16 TOPS via PyTorch Metal MPS), and Apple Neural Engine (FP16/INT8 TOPS via CoreML).
5. Logic Board & Bus Integrity Audit — Live status checks for Camera, Audio, Secure Enclave (Touch ID), Bluetooth, SMC thermal pressure, Fan RPM, Kernel Panic logs, and I2C/SPI bus mining.

---

🏗️ Modular Architecture

The diagnostic engine is organized into decoupled, independent suites designed for single-responsibility execution:

macOS-Hardware-info.py
│
├── [Data Models & Protocols]
│   ├── DiskInfo              → Hardware metadata & location descriptors
│   ├── SmartReport           → S.M.A.R.T. metrics, TBW & thermal telemetry
│   └── BenchmarkResults      → Computational throughput & hardware specs
│
├── [Suite 1: AI & Compute Benchmarks]
│   ├── benchmark_cpu()       → High-intensity FP32 matrix multiplication (GFLOPS)
│   ├── benchmark_gpu()       → FP16 Tensor GEMM via PyTorch MPS (TOPS)
│   └── benchmark_npu()       → CoreML Neural Engine MLProgram stress test (TOPS)
│
├── [Suite 2: Storage & Multi-Drive Engine]
│   ├── find_physical_disks() → Physical drive enumeration (diskutil & plistlib)
│   ├── get_disk_mount_point()→ Dynamic writable mount point resolution (/Volumes/...)
│   ├── get_smart_data()      → Multi-protocol SMART probe (NVMe, SATA, Bridge descriptors)
│   └── benchmark_disk_speed()→ Multi-threaded unbuffered I/O benchmark (fcntl.F_NOCACHE)
│
├── [Suite 3: Logic Board & Peripheral Audit]
│   ├── check_logic_board_health()  → BHM official health, IOKit mAh, SMC & Panics
│   └── check_peripherals_and_buses() → Camera, Audio, Touch ID, Bluetooth & Bus logs
│
└── [Execution Orchestrator]
    ├── select_execution_option()   → Interactive console selector
    ├── parse_arguments()           → CLI argument parser (argparse)
    └── save_report_file()          → Automated persistent .txt log generation

---

🚀 Usage & Execution Flow

The tool requires sudo privileges to communicate with IOKit, smartctl, powermetrics, and raw block storage descriptors.

# Execute with administrative privileges
sudo python3 macOS-Hardware-info.py

Interactive Console Menu

When launched without arguments, the script presents an interactive menu:

════════════════════════════════════════════════════════════════════════════════
  macOS HARDWARE INFO - Complete Hardware Diagnostics
════════════════════════════════════════════════════════════════════════════════
  Select diagnostic suite to execute:

  [1] 🚀 Run All Tests (Full Suite + Persistent .txt Report Generation)
  [2] 🧠 Performance Benchmarks (CPU, GPU & AI with PyTorch)
  [3] 💾 Storage Diagnostics (Internal & External SSDs/HDDs)
  [4] 🔬 Hardware Diagnostics (Logic Board, Peripherals & Buses)
  [5] ❌ Exit

CLI Flags (Automation & CI/CD)

For headless execution or script automation, specific flags bypass the interactive menu:

Flag	Long Argument	Description
-a	--all	Executes all test suites and generates {Mac_Model_Serial}.txt report
-i	--ai	Runs CPU, GPU, and NPU performance benchmarks
-d	--disks	Runs internal and external storage S.M.A.R.T. and I/O speed benchmarks
-w	--hardware	Runs Logic Board health, peripheral presence, and bus communication audit
-h	--help	Displays command-line help and usage options

# Examples:
sudo python3 macOS-Hardware-info.py --disks
sudo python3 macOS-Hardware-info.py --ai
sudo python3 macOS-Hardware-info.py --all

---

💾 Storage & Multi-Drive Engine

Multi-Drive & External Enclosure Support

The storage engine automatically detects and classifies both internal and external storage media (Apple Fabric NVMe, PCIe SSDs, SATA drives, USB 3.2, and Thunderbolt enclosures):

• Physical Node Discovery: Uses diskutil list -plist to parse all physical device nodes (/dev/disk0, /dev/disk1, etc.) independently of synthesized APFS containers.
• Dynamic Mount Point Resolution: Resolves active writable mount paths (/Volumes/DriveName) across partitioned and unpartitioned volumes for secondary and external drives.
• Multi-Probe S.M.A.R.T. Ingestion: Automatically attempts communication across NVMe, ATA/SATA, and SCSI translation bridge layers (-d sat, -d sntasmedia, -d sntrealtek, -d sntjmicron).
• Darwin Bridge Awareness: Accurately identifies USB mass storage limitations in macOS, displaying clear diagnostic status when hardware bridge pass-through is restricted by the operating system kernel.

Unbuffered Parallel I/O Speed Test

Measures physical drive throughput while bypassing macOS Unified Memory caching:

               [ Application Buffer (16 MB Chunks) ]
                                │
                 ┌──────────────┴──────────────┐
                 ▼ (Parallel ThreadPool Workers)▼
       Worker 1  Worker 2  Worker 3  Worker 4 ... Worker 8
                 │         │         │         │
                 └──────────────┬──────────────┘
                                │
                   fcntl.F_NOCACHE Bypass
                                │
                     [ Unified RAM (ARC) ] ── (Bypassed)
                                │
                     [ Physical NAND / Bus ]

• Cache Bypass (fcntl.F_NOCACHE): Prevents macOS Unified Memory (ARC) from caching benchmark payload blocks, ensuring direct NAND read/write measurement.
• High Queue Depth Simulation: Spawns 8 concurrent worker threads via ThreadPoolExecutor with 16 MB chunk sizes to saturate PCIe 4.0/5.0 and high-speed USB-C buses.
• Entropy Protection: Employs pseudo-random data streams (os.urandom) to prevent hardware controller compression engines from artificially inflating throughput metrics.

---

🔋 Battery & Power Subsystem

The power diagnostic module features a dual-layer architecture combining certified system metrics with low-level electrochemical data:

┌───────────────────────────────────────────────────────────┐
│              macOS Power Diagnostic Engine                │
├─────────────────────────────┬─────────────────────────────┤
│   Layer 1: Official BHM     │   Layer 2: IOKit Hardware   │
│   (SPPowerDataType XML)     │   (AppleSmartBattery Plist) │
├─────────────────────────────┼─────────────────────────────┤
│ • Maximum Capacity (98%)    │ • Design Capacity (8579 mAh)│
│ • Health Condition (Normal) │ • Nominal Charge (8305 mAh) │
│ • Calibrated System Level   │ • Raw Max Charge (8061 mAh) │
│                             │ • Cycle Count & Voltage     │
│                             │ • Real-time Cell Temp (°C)  │
└─────────────────────────────┴─────────────────────────────┘

1. Official Battery Health Management (BHM): Retrieves certified maximum capacity percentages matching Apple System Settings 1:1 via structured XML parsing.
2. Electrochemical IOKit Registry: Extracts factory design capacity (DesignCapacity), temperature-compensated nominal capacity (NominalChargeCapacity), instantaneous fuel gauge registers (AppleRawMaxCapacity), cycle counts, cell voltage, and thermistor readings.

---

🧠 AI & Silicon Benchmarks

Designed to measure true compute roofline limits across all execution units on Apple Silicon and Intel hardware:

Target	Framework / Engine	Workload	Metric
CPU	NumPy (Apple Accelerate BLAS)	2048×2048 FP32 Matrix Multiplication	GFLOPS (FP32)
GPU	PyTorch (Metal Performance Shaders)	4096×4096 FP16 Tensor GEMM	TOPS (FP16)
NPU	CoreML (Apple Neural Engine)	5-layer Deep Convolutional Network	TOPS (FP16 & INT8)

• Deterministic Hardware Stress: GEMM (General Matrix Multiply) operations maintain an arithmetic intensity $> 1300\text{ FLOPs/Byte}$, ensuring benchmarks remain compute-bound without memory-bandwidth throttling.
• Apple Silicon Metal Acceleration: Directly utilizes torch.mps with hardware synchronization barriers (torch.mps.synchronize()) for consistent throughput measurement.

---

📦 Installation & Setup

Prerequisites

Requirement	Supported Versions	Verification
macOS	11.0+ (Big Sur to Sequoia)	sw_vers
Python	3.10 or 3.11	python3 --version
System Tool	smartmontools	smartctl --version

Quick Start

# 1. Clone repository
git clone https://github.com/CodeGeekR/macOS-hardware-info.git
cd macOS-hardware-info

# 2. Install system dependency
brew install smartmontools

# 3. Create virtual environment
python3.11 -m venv venv
source venv/bin/activate

# 4. Install Python dependencies
pip install --upgrade pip setuptools wheel
pip install -r requirements.txt

# 5. Run diagnostics suite
sudo python3 macOS-Hardware-info.py

---

📄 License & Author

Distributed under the MIT License. See LICENSE for details.

Copyright (c) 2024 samuraidev

Author: samuraidev
Portfolio: arquitectodesoftware.co

⭐ Star this project if you find it useful!

Built with precision for the macOS developer community