SNIA Developer Conference September 15-17, 2025 | Santa Clara, CA
Abstract
Computational storage (CS) transforms how we handle data, and this paper takes a deep dive into how compression impacts performance and Total Cost of Ownership (TCO) for storage systems. Computational storage brings compute closer to the data and accelerates common data processing workloads, including data reduction techniques like compression to free up valuable CPU resources. Compression impact on TCO is a huge lever. It can help us store more with the same space and reduce the total cost - but compression can slow things down and put more work on the CPU. We turn to computational storage to seamlessly offload the compression. We will look at a case study with Ceph, enabling compression on a single node cluster and comparing it with computational storage. We will use the SNIA Storage TCO model to show the impact of CS on storage costs.