Computational Storage offers near-data acceleration, and it is gaining popularity with recent commercialization and standardization efforts. In this talk, we present how Computational Storage can be used to scale the performance of a key-value storage engine and deep learning training workloads. We propose a new key-value storage engine, named RETINA, where Computational Storage, Samsung SmartSSD, accelerates its data processing and user-defined processing pipelines. RETINA adopts a cross-layered approach between the host CPU and Computational Storage to leverage the host’s flexibility, fine-grained control and Computational Storage’s near-data acceleration. We show how RETINA can improve the performance and efficiency of key-value storage and deep learning workloads.