Benchmarking the Security Protocol and Data Model (SPDM) for component authentication
Authors:
Renan C. A. Alves,
Bruno C. Albertini,
Marcos A. Simplicio Jr
Abstract:
Efforts to secure computing systems via software traditionally focus on the operating system and application levels. In contrast, the Security Protocol and Data Model (SPDM) tackles firmware level security challenges, which are much harder (if at all possible) to detect with regular protection software. SPDM includes key features like enabling peripheral authentication, authenticated hardware meas…
▽ More
Efforts to secure computing systems via software traditionally focus on the operating system and application levels. In contrast, the Security Protocol and Data Model (SPDM) tackles firmware level security challenges, which are much harder (if at all possible) to detect with regular protection software. SPDM includes key features like enabling peripheral authentication, authenticated hardware measurements retrieval, and secure session establishment. Since SPDM is a relatively recent proposal, there is a lack of studies evaluating its performance impact on real-world applications. In this article, we address this gap by: (1) implementing the protocol on a simple virtual device, and then investigating the overhead introduced by each SDPM message; and (2) creating an SPDM-capable virtual hard drive based on VirtIO, and comparing the resulting read/write performance with a regular, unsecured implementation. Our results suggest that SPDM bootstrap time takes the order of tens of milliseconds, while the toll of introducing SPDM on hard drive communication highly depends on specific workload patterns. For example, for mixed random read/write operations, the slowdown is negligible in comparison to the baseline unsecured setup. Conversely, for sequential read or write operations, the data encryption process becomes the bottleneck, reducing the performance indicators by several orders of magnitude.
△ Less
Submitted 12 July, 2023;
originally announced July 2023.
