SNIA Developer Conference September 15-17, 2025 | Santa Clara, CA
Welcome and Opening Remarks
Data bytes stored continues to grow at about 40% annually. This trend now exceeds the device capacity growth rate of all existing commercial scale media types including HDD, Flash, Tape and Optical, and the gap between growth rates is about 20%. That implies that the datacenter footprint for storage will be approximately doubling every 3.5 years just to keep up. However, the roadmaps for ongoing density improvement makes the situation much more stark. Past 2030, growth in device capacities may slow substantially leading to a need for 40x or more datacenter space and power by 2040 to keep up with data growth. Although the methods we have used to store data with magnetized materials or corralling electrons are true technological wonders, it is becoming apparent that if we don't want to impinge data growth we may need a substantial paradigm shift in storage technology. Molecular storage is the panacea of storage density and DNA is the leading contender for the championship of storage density, but we will also need to invest in technologies that allow for high speed molecular storage. This is going to be a heavy lift, but if we want to intercept the coming storage capacity crunch we need to start work now.
Software-Enabled Flash™ technology is being developed to allow storage developers to maximize the value of flash deployed in their data centers. It makes flash storage software-defined by redefining the relationship between the host and solid-state storage. Purpose-built, media-centric hardware focused on cloud-scale requirements, combined with an open source API and SDK, give developers tools to define latency outcomes, tenant isolation, control flash background processes, and build their own application-centric storage protocols. The Software-Enabled Flash Project, under the Linux Foundation, was formed to advance this new technology and encourage collaboration and innovation in a vendor neutral forum. Learn about this new open source project, details of the Software-Enabled Flash technology, and the progress of groundbreaking engineering sample drives implementing this new, developer-focused paradigm.
Los Alamos is working to revolutionize how scientific data management is done, moving from large Petabyte sized files generated periodically by extreme scale simulations to a record and column based approach. Along the journey, the NVME Computational Storage efforts became a strategic way to help accomplish this revolution. Los Alamos has been working on a series of proof of concepts with a set of data storage industry partners and these partnerships have proven to be the key to success. This talk will summarize the three proof of concept applications of Computation Storage and some of the industry partnership projects that have helped pave the way for LANL and hopefully industry to new approaches to large scale data management.
The exponential growth in data volume and traffic across the global internet infrastructure is motivating the exploration of new architectures for data centers to achieve the demand for better performance, efficiency and total cost of ownership (TCO). One promising trend is to segregate the functional components of compute, memory, storage, and networking into pools and organize them as needed according to the specific needs of different workloads. This disaggregation and composability will take us beyond the classic architecture of servers as the unit of computing. Backed by a broad consortium of hyperscalers, equipment OEMs, chipmakers and IP vendors, Compute Express Link (CXL) is a new enabling technology for interconnecting computing resources that embraces existing competing technologies such as Gen-Z and OpenCAPI. Currently in Gen 3.0, CXL enables high-speed, low-latency links with memory cache coherency between processors, accelerators, NICs, memory, and storage. In this keynote, Dr. Ki is going to present the opportunities and challenges that exist across hardware and software, with a focus on Samsung's CXL memory technologies such as CXL memory expanders and memory semantic SSDs. In particular, he will share the direction Samsung wants to move forward with the storage community and suggest various collaboration opportunities.
Given LANL simulations can generate a Petabyte of data per time step with thousands to tens of thousands of time steps, data gravity is a huge concern at the lab. Performing analytics on this data to find and understand interesting features on simulation output is extremely expensive requiring movement of Petabytes of data and a data analytics platform with enough memory, not storage, to hold a full timestep (Petabyte). Often features of interest exist in far less data than the total data, like tracking the front of a shockwave going through a material, where you only care about the region at the very front of the shock wave. Sometimes the actual interesting feature is captured in several orders of magnitude less data the overall simulation output. We have been keenly interested in computational storage techniques to reduce the data as close to the storage as possible, making the amount of data that has to be moved and the size of the analytics platform many times if not orders of magnitude smaller. LANL and its many partners have explored many avenues including row, column, object, file and even block based approaches. This talk will provide an overview of this exploration and its future direction.