Creating a Framework for Computational Storage Applications

Computational Storage (putting compute power where the data resides) is an intriguing approach with obvious benefits that could change the way computer systems handle large datasets. However, for its use to spread beyond a few highly advanced customers with the resources to do significant internal development, it must have a framework that provides basic tools, methods, and functionality. The framework must offer infrastructure that enables security, allows for various types of data to be handled across the entire environment, and provides interfaces for a variety of hardware (such as CPUs,FPGAs, GPUs, or accelerators), development environments, and management tools. It must both support near term products getting to market and set a foundation for solving scaling and other long-term issues. It will allow typical developers with limited resources and expertise to utilize computational storage in an ever wider range of applications. The result will be much larger markets, a bigger and stronger ecosystem, and tremendous advances in the overall technology.