Embedded & MCU Explainer How To IoT Hardwares Microcontrollers Raspberry Pi Sensor & Devices Tutorials/DIY Useful Stuff

PCIe 3.0 to Dual M.2 HAT for Raspberry Pi 5

5th February 2025 Harshvardhan Mishra 0 Comments 5 min read

Introduction

The PCIe 3.0 to Dual M.2 HAT by Seeed Studio is an advanced expansion board designed to enhance the functionality of the Raspberry Pi 5. This HAT provides support for dual M.2 NVMe SSDs, enabling high-speed storage expansion and PCIe-based peripheral connectivity. It is ideal for users looking to improve data transfer speeds, storage performance, and expandability on their Raspberry Pi 5.

This guide will cover the specifications, features, installation, use cases, and performance benchmarks of the PCIe 3.0 to Dual M.2 HAT, providing a comprehensive resource for Raspberry Pi enthusiasts.

Specifications

Below is a detailed specifications table for the PCIe 3.0 to Dual M.2 HAT:

Specification	Details
Interface	PCIe 3.0 x1
M.2 Slots	2x M.2 M-key slots (NVMe SSD support)
Compatibility	Raspberry Pi 5
Supported SSDs	M.2 NVMe (2242/2280 sizes)
Power Supply	Powered via Raspberry Pi 5
Data Transfer Speed	Up to 8 GT/s
Mounting Options	Screw holes for stable installation

Key Features

1. Dual NVMe SSD Support

The HAT allows users to install two M.2 NVMe SSDs, significantly boosting the storage capacity and read/write speeds of the Raspberry Pi 5.

2. PCIe 3.0 High-Speed Interface

With support for PCIe 3.0 x1, the board achieves high data transfer speeds, making it ideal for data-intensive applications, high-speed computing, and fast boot times.

3. Plug-and-Play Installation

Designed for the Raspberry Pi 5, the HAT features easy installation with a secure PCIe connection, requiring minimal hardware setup.

4. Compact and Efficient Design

The board maintains a low-profile and compact design, ensuring optimal space management while keeping the Raspberry Pi 5 lightweight.

5. Extended Storage & Performance Boost

With the ability to install dual NVMe SSDs, the HAT significantly improves data access speeds, making it perfect for server applications, database management, and AI workloads.

Applications

High-Speed Storage Expansion – Use NVMe SSDs for fast boot, application loading, and data storage.
Edge Computing & AI Workloads – Enhances Raspberry Pi 5 for machine learning, AI inference, and analytics.
Personal Cloud & NAS Setup – Build a low-cost NAS or cloud storage server.
High-Performance Computing (HPC) – Ideal for data-intensive projects requiring high-speed read/write operations.
Media Center & Game Emulation – Provides faster storage for Kodi, Plex, and gaming emulators.

Setting Up the PCIe 3.0 to Dual M.2 HAT

Step 1: Unboxing and Hardware Setup

Ensure you have the PCIe 3.0 to Dual M.2 HAT, Raspberry Pi 5, and M.2 NVMe SSDs.
Align and connect the HAT to the PCIe connector on the Raspberry Pi 5.
Secure the NVMe SSDs using the provided mounting screws.

Step 2: Configuring the Raspberry Pi 5

Power on the Raspberry Pi 5 and update the system:
```
sudo apt update && sudo apt upgrade
```
Install necessary PCIe drivers:
```
sudo apt install nvme-cli
```
Check if the SSDs are detected:
```
lsblk
```

Step 3: Formatting and Mounting SSDs

Format the SSD for use:
```
sudo mkfs.ext4 /dev/nvme0n1
```
Create a mount point:
```
sudo mkdir /mnt/ssd1
```
Mount the SSD:
```
sudo mount /dev/nvme0n1 /mnt/ssd1
```
To auto-mount on boot, add the following to /etc/fstab:
```
/dev/nvme0n1 /mnt/ssd1 ext4 defaults 0 0
```

Performance Testing & Optimization

Step 1: Benchmarking NVMe SSDs

Install the benchmarking tool:
```
sudo apt install fio
```

Run a read/write test:

fio --name=test --size=1G --rw=readwrite --bs=4k --numjobs=4 --time_based --runtime=30s --group_reporting

Check SSD health:
```
sudo nvme smart-log /dev/nvme0n1
```

Step 2: Optimizing NVMe Performance

Enable TRIM for SSD longevity:
```
sudo systemctl enable fstrim.timer
```

Adjust I/O scheduler for better performance:

echo "none" | sudo tee /sys/block/nvme0n1/queue/scheduler

Why Choose PCIe 3.0 to Dual M.2 HAT?

Feature	Benefit
High-Speed Storage	Supports up to PCIe 3.0 speeds for faster data transfer
Dual M.2 NVMe Support	Enables massive storage expansion for Raspberry Pi 5
Plug-and-Play Design	Easy installation with minimal configuration required
Ideal for Power Users	Boosts performance for AI, NAS, and edge computing
Compact & Reliable	Designed for low power consumption and efficient space management

Step-by-Step Guide: Setting Up a NAS or AI Processing Unit with PCIe 3.0 to Dual M.2 HAT

Part 1: Setting Up a NAS with PCIe 3.0 to Dual M.2 HAT

Step 1: Hardware Setup

Connect the PCIe 3.0 to Dual M.2 HAT to the Raspberry Pi 5.
Install one or two NVMe SSDs securely in the provided M.2 slots.
Attach necessary peripherals (Ethernet, USB keyboard/mouse, and display).
Power on the Raspberry Pi 5.

Step 2: Software Installation

Update the system:
```
sudo apt update && sudo apt upgrade
```

Install OpenMediaVault (OMV) for NAS management:

wget -O - https://github.com/OpenMediaVault-Plugin-Developers/installScript/raw/master/install | sudo bash

After installation, access OMV’s web interface at http://<your-raspberry-pi-ip>/.

Step 3: Configure Storage and Shares

Under Storage > Disks, verify that your NVMe SSDs are detected.
Go to File Systems, create and mount a new file system for storage.
Under Shared Folders, create folders to store data.
Enable SMB/NFS under Services > SMB/NFS for network access.
Access your NAS from other devices using the assigned IP address.

Part 2: Setting Up an AI Processing Unit

Step 1: Preparing the AI Environment

Install necessary AI libraries:

sudo apt install python3-pip
pip3 install torch torchvision torchaudio

Install OpenCV for image processing:
```
sudo apt install python3-opencv
```
Set up TensorFlow and Keras:
```
pip3 install tensorflow keras
```

Step 2: Optimizing AI Processing on NVMe SSD

Configure the swap file to use NVMe SSD for caching:

sudo fallocate -l 4G /mnt/nvme0n1/swapfile
sudo chmod 600 /mnt/nvme0n1/swapfile
sudo mkswap /mnt/nvme0n1/swapfile
sudo swapon /mnt/nvme0n1/swapfile

Enable hardware acceleration:
```
sudo apt install libatlas-base-dev
```

Run a sample AI inference:

python3 -c "import torch; print(torch.cuda.is_available())"

Step 3: Running AI Models

Clone a sample AI project:

git clone https://github.com/pjreddie/darknet
cd darknet
make

Run object detection using YOLO:

./darknet detect cfg/yolov3.cfg yolov3.weights data/dog.jpg

Deploy AI models using TensorFlow:
```
python3 my_model.py
```

Conclusion

The PCIe 3.0 to Dual M.2 HAT by Seeed Studio is an essential upgrade for Raspberry Pi 5 users seeking high-speed storage expansion and PCIe-based connectivity. Whether you’re setting up a NAS, AI workstation, or media center, this expansion board ensures fast, reliable, and scalable storage solutions.