Where Can Computational Storage Make a Difference?

This week in Las Vegas, VMWare held the VMworld 2018 summit. While a lot of the discussion was about moving applications to the cloud, there was equally as much discussion about edge computing, the Internet of Things (IoT), and petabyte-scale real-time analytics. While computational storage can’t help with the first of these topics (at least not until AWS, Microsoft Azure, or Google Cloud starts using our products), we can definitely help with the last two of these. One of the places where computational storage shines is read-centric parallel analytics (e.g., lots of parallel operations on a single dataset). Last month we talked about how we accelerated the Facebook Artificial Intelligence Similarity Search (FAISS) algorithm using our computational storage devices; over the next few months we will also run benchmarks against other applications that we can have a positive impact on. We also expect to have more to say soon on IoT as we work with Microsoft Research. Stay tuned!