More from Cigent
Blog
•
16 minute read
•
October 6, 2026
Hardware vs. software encryption is an important distinction for Authorizing Officials and program managers responsible for protecting classified data at rest. Both approaches make stored information unreadable without the cryptographic keys required to decrypt it, but they differ in where encryption occurs and how the protection is implemented.
Understanding that difference is increasingly important as classified data moves beyond traditional workstations onto tactical servers, edge-computing systems, autonomous platforms, and other devices operating outside traditional secure environments.
This article explains how hardware and software encryption work, the advantages and considerations of each approach, why encryption alone does not constitute a complete classified DAR architecture, and how NSA’s Commercial Solutions for Classified framework uses independent encryption layers to protect classified information.
Full drive encryption protects information stored on a drive by transforming readable data, called plaintext, into unreadable ciphertext. The appropriate cryptographic key is required to transform that information back into usable data.
For full drive encryption, this process occurs continuously. Information written to storage is encrypted, and information retrieved by an authorized system is decrypted.
Hardware and software encryption accomplish this basic objective differently. The primary distinction is where the cryptographic process occurs and how the encryption capability is implemented.
Hardware encryption performs the cryptographic function within the storage device itself, typically through a self-encrypting solid-state drive.
The drive contains dedicated hardware and firmware responsible for encrypting and decrypting information. Encryption therefore occurs within the storage device rather than relying primarily on the host operating system to perform the cryptographic operation.
This architecture provides several potential advantages:
These characteristics can make hardware encryption a useful foundation for protecting data at rest.
However, an encrypted drive is not necessarily a complete protected-storage architecture. Programs also need to consider what controls access to the encrypted drive and under what conditions its protected information becomes accessible.
Software encryption performs the cryptographic function through software running on the computing platform.
Rather than relying on the storage device itself to perform the encryption, software encrypts information before it is written to storage and decrypts it when authorized access is required.
One of the primary advantages of software encryption is deployment flexibility. It can potentially be deployed across existing systems without replacing their physical storage devices.
That can be particularly useful for organizations managing large installed bases of laptops, workstations, servers, tactical computers, and other systems already operating in the field.
Replacing storage hardware can involve opening systems, installing new devices, reimaging equipment, managing logistics, and evaluating configuration impacts. A software encryption layer can potentially extend protection across existing systems through software deployment.
Software encryption can also provide greater independence from a specific storage device and support standardization across heterogeneous hardware environments.
Because software encryption operates within the computing platform, however, its security also depends on how the software, operating environment, authentication, cryptographic keys, and platform are configured and protected.
Hardware and software encryption should not be viewed simply as competing technologies. Each approach has characteristics that can make it appropriate for different systems, architectures, and deployment requirements.
| Consideration | Hardware Encryption | Software Encryption |
|---|---|---|
| Where encryption occurs | Within the storage device | Through software running on the computing platform |
| Storage dependency | Encryption capability is integrated into the drive | Can provide greater independence from a specific storage device |
| Existing systems | May require installation or replacement of storage hardware | Can potentially be deployed without replacing existing storage |
| Encryption processing | Performed by capabilities within the storage device | Performed through software on the computing platform |
| Host environment | Cryptographic processing can be isolated from portions of the host environment | Operates more closely with the host operating environment |
| Deployment flexibility | Well suited to systems designed around compatible encrypted storage | Can support deployment across heterogeneous and already-fielded systems |
Neither approach is universally better. The appropriate choice depends on the system, mission, operating environment, existing hardware, applicable requirements, and broader security architecture.
Encryption is a security mechanism. It is not, by itself, a complete data-protection architecture.
A useful analogy is a safe. Encryption creates the safe that protects the information inside. Authentication provides the lock controlling whether access to the protected information is permitted.
This distinction becomes especially important when protecting classified systems against physical loss or capture.
Consider a laptop, tactical server, or autonomous platform containing an encrypted SSD. Programs still need to determine what occurs when the complete system is physically possessed by an unauthorized party.
The relevant question is not simply whether the drive is encrypted.
Programs should also ask what prevents someone who possesses the entire device from accessing the protected information.
Encryption is part of that answer. Authentication, cryptographic key protection, system configuration, access controls, and the overall security architecture complete it.
Authentication controls whether access to protected storage is permitted before the information becomes accessible.
An encrypted drive can be viewed as the safe, while authentication provides the lock controlling access to it.
This is why an encrypted drive and its pre-boot authentication should be considered together when discussing the hardware protection layer. Pre-boot authentication is not a separate encryption layer.
Programs should evaluate what authentication occurs before protected storage becomes accessible, how authentication information is protected, and how the system behaves when it is disconnected from enterprise services or outside authorized physical control.
FIPS 140-3 establishes security requirements for cryptographic modules. Validation provides assurance that specified cryptographic functionality has been tested against defined requirements.
That validation is important, but it applies to the cryptographic module rather than establishing that the complete system implements an architecture appropriate for protecting classified data at rest.
A system can therefore contain FIPS-validated cryptography while still requiring evaluation of authentication, key protection, system configuration, independent protection layers, and other security functions.
For an Authorizing Official or program manager, FIPS validation answers important questions about the cryptographic module. It does not, by itself, answer the broader operational question of what happens to classified information when the complete system is lost, stolen, or captured.
NSA’s Commercial Solutions for Classified program enables commercial technologies to be used within defined architectures for protecting classified National Security Systems data.
The current Data-at-Rest Capability Package is Version 5.1.0, dated March 2026. It provides high-level reference designs and configuration requirements for protecting classified data at rest.
The fundamental concept is defense in depth through two independent encryption layers.
NSA’s CSfC Customer Handbook states that the DAR Capability Package enables customers to implement two independent layers of encryption for protecting stored information while the system is powered off or remains in an unauthenticated state.
The specific technologies used for those layers depend on the applicable CSfC DAR solution design.
One CSfC DAR approach combines hardware full drive encryption with independent software full drive encryption.
For applicable Cigent architectures, the layers can be understood as:
The encrypted drive and its pre-boot authentication form the hardware layer. Pre-boot authentication should not be counted as a separate third layer.
The value of the architecture comes from independence. If one protection layer is compromised, another independent cryptographic layer remains between an unauthorized party and the classified information.
No. Hardware encryption plus software encryption is one CSfC DAR solution design, but it is not the only design supported by the current Capability Package.
Data-at-Rest Capability Package v5.1.0 introduced a software full drive encryption/software full drive encryption solution design.
This distinction is important. The governing CSfC principle is not simply that every system must use one hardware encryption product and one software encryption product.
The architecture requires two appropriately configured, independent encryption layers implemented according to the applicable solution design and Capability Package requirements.
Two independent encryption layers provide defense in depth so that protection of classified information does not depend entirely on one cryptographic implementation.
Consider a classified tactical computer using only a hardware-encrypted SSD. The drive may provide strong encryption, but the overall architecture still depends on the controls protecting access to that layer.
Adding an independent software encryption layer creates another cryptographic barrier around the classified information.
Compromising one layer therefore does not automatically mean that the second layer is also compromised.
The independence between those layers is an important part of the CSfC architecture.
Hardware versus software encryption should not be treated as a simple either-or technology decision.
Hardware encryption can provide purpose-built cryptographic protection integrated directly into storage. Software encryption can provide flexible protection across heterogeneous platforms and may be useful for extending protection to systems already deployed in the field.
For classified systems, however, the discussion needs to extend beyond individual encryption technologies and address the complete DAR architecture.
Authorizing Officials and program managers should ask:
| Question | Why It Matters |
|---|---|
| Where does encryption occur? | Determines whether cryptographic processing occurs within storage hardware or through software on the computing platform. |
| Is existing hardware being retained? | Software encryption may support deployment without replacing storage devices. |
| How is access authenticated? | Encryption needs controls governing when protected information becomes accessible. |
| How are keys protected? | Cryptographic protection depends on appropriate handling and protection of associated keys. |
| What happens during physical capture? | Classified DAR architecture should account for an unauthorized party possessing the complete endpoint. |
| Are the layers independent? | CSfC DAR relies on independent encryption layers rather than duplicating the same protection mechanism. |
| Which solution design applies? | The current DAR Capability Package supports multiple solution designs. |
NSA currently publishes the Data-at-Rest Capability Package Version 5.1.0, dated March 2026. The Capability Package provides high-level reference designs and corresponding configuration requirements for protecting classified data at rest using commercial technologies.
NSA’s CSfC Customer Handbook states that the DAR Capability Package enables customers to implement two independent layers of encryption to protect stored information on an end-user device or DAR-protected system while it is powered off or in an unauthenticated state.
Version 5.1.0 also introduced a software full drive encryption/software full drive encryption solution design, expanding the available architectures beyond designs combining hardware and software encryption.
Programs should verify the current Capability Package, applicable solution design, configuration requirements, and component status against NSA’s live public resources during architecture and procurement decisions.
Hardware and software encryption both provide important mechanisms for protecting data at rest, but they do so differently. Hardware encryption integrates cryptographic protection into the storage device, while software encryption can provide flexible protection across existing and heterogeneous computing environments.
For classified information, the objective is not simply to choose between hardware and software encryption or confirm that a drive is encrypted. Programs need to evaluate authentication, cryptographic key protection, physical-capture scenarios, independence between protection layers, and the complete architecture required for the system and information being protected.
For Authorizing Officials and program managers evaluating classified DAR requirements, take the CSfC DAR Readiness Assessment to identify architecture considerations and areas that may require further review.
Take the CSfC DAR Readiness Assessment
Full drive encryption protects information stored on a drive by converting readable plaintext into ciphertext that requires the appropriate cryptographic keys to decrypt. Encryption occurs as information is written to storage, while authorized access allows the information to be decrypted when needed. Hardware and software approaches perform this function differently.
Hardware encryption performs cryptographic operations within the storage device using capabilities integrated into the drive. Software encryption performs encryption through software running on the computing platform. Both protect stored information, but they differ in implementation, hardware dependency, deployment flexibility, processing location, and their relationship with the host operating environment.
Hardware encryption integrates cryptographic protection directly into the storage device. Encryption processing occurs within the drive, and removing the SSD from its host system does not remove its encryption. These characteristics can provide a useful foundation for DAR protection when combined with appropriate authentication, key protection, configuration, and architecture.
Software encryption can provide flexibility across existing and heterogeneous computing environments. It may be deployed without replacing physical storage devices, which can be useful for systems already operating in the field. Software encryption can also provide greater independence from a specific drive, depending on the implementation and applicable security requirements.
Neither approach is universally better. Hardware encryption provides purpose-built protection integrated into storage, while software encryption provides deployment flexibility and can support existing systems. The appropriate approach depends on the mission, system architecture, hardware, operating environment, applicable security requirements, and the role each technology performs within the complete DAR architecture.
An encrypted SSD provides an important security function, but programs must also evaluate authentication, cryptographic key protection, startup behavior, system configuration, and physical-capture scenarios. For applicable classified systems, DAR requirements address the complete protection architecture rather than treating the presence of an encrypted drive as the security outcome.
No. FIPS 140-3 validation provides assurance about specified cryptographic functionality within a cryptographic module. It does not by itself establish that a complete device implements an architecture appropriate for classified DAR. Programs must also evaluate authentication, key protection, configuration, independent layers, and applicable NSA requirements for the complete solution.
NSA’s CSfC DAR architecture uses two independent layers of encryption. Depending on the applicable solution design, those layers can use different combinations of encryption technologies. In a hardware and software architecture, an encrypted drive with pre-boot authentication forms one layer and independent software full drive encryption provides the second layer.
Not in every solution design. Data-at-Rest Capability Package v5.1.0 introduced a software full drive encryption/software full drive encryption design. The governing principle is two appropriately configured independent encryption layers that meet the applicable Capability Package requirements, rather than a universal requirement to combine hardware and software encryption.
An Authorizing Official should evaluate where classified information is stored, how it is encrypted, what authentication controls access, how cryptographic keys are protected, what happens during physical capture, whether independent protection layers remain effective, which CSfC DAR solution design applies, and whether the complete architecture meets applicable requirements.
Brett Hansen is Chief Executive Officer of Cigent, a cybersecurity company advancing the protection of sensitive and classified data. A technology executive with more than 30 years of experience, he has held senior leadership roles across cybersecurity, enterprise software, and SaaS. He spent more than a decade at Dell Technologies, where he led the Client Software organization and built Dell Data Security, and began his career at IBM in software leadership roles spanning IT management, development, and security. Under Hansen's leadership, Cigent has achieved more than 100 percent year-over-year growth while emerging as the leader in closing a critical national security gap: protecting data at rest wherever the mission operates. Cigent offers the industry's broadest portfolio of NSA-listed CSfC Data-at-Rest technologies and continues to advance data protection through quantum-resistant cryptography, secure data sanitization, and layered hardware and software security. Hansen is committed to translating innovation into resilient, practical security for government, defense, and enterprise organizations.
