Podman/Docker setup for running IPED with GPU acceleration. Includes audio transcription (Whisper), face recognition (dlib), OCR, and image processing—all optimized for NVIDIA CUDA.

# Fedora/RHEL
sudo dnf install podman podman-compose

# Ubuntu/Debian
sudo apt install podman podman-compose

# Arch
sudo pacman -S podman podman-compose

# Install NVIDIA Container Toolkit for GPU support
# https://docs.nvidia.com/datacenter/cloud-native/container-toolkit/

podman build docker/ -t iped-cuda:4.3.0

./startIped-cli.sh process \
  --evidence /path/to/evidence.E01 \
  --hashes-db /path/to/hashdb \
  --output my_case

# Multiple evidence files
./startIped-cli.sh process \
  --evidence /data/phone.E01 \
  --evidence /data/disk.dd \
  --hashes-db /db/hashes \
  --output my_case

./startIped-cli.sh analyze --case-name my_case

./startIped-cli.sh process [options]

Required:

• -e, --evidence PATH - Evidence file/directory (repeatable)
• -d, --hashes-db PATH - Hash database directory
• -o, --output NAME - Case output name

Optional:

• -t, --threads NUM - Processing threads (default: half available cores)
• -m, --memory SIZE - Java heap size (default: 2/3 physical RAM)
• -c, --config PATH - Custom config directory
• --continue - Resume interrupted case
• --no-gpu - Disable GPU acceleration
• --nogui - Headless mode

./startIped-cli.sh analyze --case-name NAME

./startIped-cli.sh list

./startIped-cli.sh clean

To configure the behaviour or profiles of IPED, just modfiy the files in conf/ before starting a new case. You can additionally edit IPEDConfig.txt or LocalConfig.txt!

All configuration files will be mounted into the container at runtime. So theres no need to rebuild the image ^^

IPED-cuda/
├── startIped-cli.sh                        # CLI entry point
├── docker/
│   ├── Dockerfile                          # Container definition
│   ├── docker-compose.template.yml         # Processing templatee
│   └── ewf.cpp.patch                       # Sleuthkit patch
├── conf/                                   # Configuration files
├── results/                                # Output (case results)
├── logs/                                   # Processing logs
└── ipedtmp/                                # Temporary files

podman run --rm --device nvidia.com/gpu=all ubuntu nvidia-smi

# If CDI error:
sudo nvidia-modprobe -u -c=0

If any of the above commands dont work, iped-cuda will never work either! So ask the internet or LLM's for help until you get that fixed basically!

• Reduce threads: --threads 4
• Reduce heap: --memory 32G
• Process in smaller batches
• Close other applications

• Check GPU usage: compare with --no-gpu flag

Licensed under GPLv3. See LICENSE.

Related projects:

• IPED (GPLv3+)
• IPED Docker
• Whisper (MIT)
• dlib (Boost License 1.0)

Docs:

• IPED Wiki
• NVIDIA Container Toolkit