Engineering updates & case studies

Latest customization work, new capabilities, and lessons from real device deployments. Click any article to read the full story.

HORM vs. standard boot: when to use Hibernate Once, Resume Many

Swayline Engineering · 6 min read · Updated 2026

Hibernate Once, Resume Many (HORM) is a Windows Embedded / IoT feature that stores a single hibernation image on disk and restores it on every cold boot. Instead of a full OS boot sequence, the system jumps straight into the saved memory state in under 5 seconds.

When HORM makes sense

HORM is ideal for devices that always run the same application in the same state: kiosks, digital signage, point-of-sale terminals, and industrial control panels. Every boot returns to an identical, known-good state - no configuration drift, no surprise updates, no user data left behind.

  • Speed: a 30-60 second cold boot compresses to under 5 seconds.
  • Consistency: every boot is identical - the same processes, the same memory layout, the same application state.
  • Simplicity: no user data to manage, no session cleanup, no post-boot scripting.

When standard boot is better

Not every device benefits from HORM. If the device needs to accept user data, update applications in the field, or support different workflows between reboots, a standard boot is more flexible.

  • Data-writing applications: HORM reverts the entire system volume on reboot. If the application needs to save data between sessions, you must redirect writes to an unprotected partition or use UWF volume exemptions.
  • Firmware and driver updates: a HORM image is frozen at deployment. Updating drivers or firmware requires re-capturing the hibernation image, which adds deployment complexity.
  • Multi-user or multi-role devices: HORM is a single-state solution. Devices that serve different users or switch between modes need a full boot to initialize each session correctly.

Combining HORM with UWF

The most common deployment pattern is HORM + UWF (Unified Write Filter). UWF intercepts all writes to the system volume and discards them on reboot. HORM ensures the boot is fast and consistent. Together they create a sealed, self-restoring system.

The key challenge is application writes. If the application needs to persist data (logs, configuration, database files), those writes must go to an unprotected location:

  • UWF volume exemptions: specific folders or partitions can be excluded from the write filter, allowing persistent writes while the rest of the system remains sealed.
  • Separate data partition: a dedicated D: drive (or similar) outside the UWF-protected volume handles all application data.
  • HORM re-capture: if the application state itself needs to change (new version, new configuration), the HORM image must be re-captured and redeployed.

Decision framework

Use this simple framework when deciding between HORM and standard boot:

  1. Does the device run a single application? If yes, HORM is a strong candidate.
  2. Does the application write data that must survive a reboot? If yes, plan for UWF exemptions or a separate data partition.
  3. Does the device need field-updatable firmware or drivers? If yes, standard boot is simpler to maintain.
  4. Is boot speed critical? If yes, HORM delivers sub-5-second boots consistently.

Need help deciding whether HORM is right for your device? Send us your hardware spec and use case.

Contact Engineering
← PreviousWindows 10 Pro vs Windows 10 IoT Enterprise: full feature comparisonNext →BitLocker, TPM and Secure Boot for medical-device deployments