High Performance Object Storage (HPOS) stack enables disaggregated storage over multiple SmartSSDs (Samsung’s computational storage drives). A data centric solution abstracting which computational function runs in which accelerator layer (i.e. CPU, GPU, XPU or SmartSSD) with ease of scale. Samsung’s HPOS solution provides reference solution for Video AI applications requiring in-storage video pre-processing. And for Data Analytics use case providing high performance for unstructured/object data in data lake.
• Data intensive real time anlytics and AI
• Surveillance, Smart City, near real-time streaming
• Data Center & IIoT Cyber Security
• Accelerate Data Lake Queries
• Low-latency ML preprocessing
• Better Throughput per watt
• Less Network traffic
• Lower TCO due to reduced CPU and network traffic
Samsung has developed DSS, a rack-scalable, very high read-bandwidth-optimized, Amazon S3-compatible object storage solution.
It utilizes a disaggregated architecture, enabling independent scaling of storage and compute. DSS is designed to make the most of system-level design, and Samsung’s best SSDs, to get the maximum performance while minimizing OPEX costs.
• Large-scale, high-throughput training
• Image Analytics
• Audio/Video AI
• High throughput storage access- 3x better than NFS
• 20% better than GPU accelerated Leading Enterprise Filesystem
• True Scalable solution
• Lower TCO with less storage nodes serving more clients
2021 OCP Global Summit
High Performance and Hardware Acceleration – With exponential data generation rate, specifically in applications like Deep learning, AI the demand for storage with high-bandwidth and great scalability that supports unstructured data format is increasing. To fulfill this need Samsung proposes DSS storage solution, which implements object Key-Value API on top NVMeOF SSD. The support of storage remote access protocols facilitates the disaggregation. Therefore, storage can be easily scaled. Besides object storage support and scalability, our architecture can provision the bandwidth demands for each application on each client server. This paper introduces our DSS Storage systems that support high-bandwidth per capacity for object-format data with an effortlessly scale-up feature. DSS uses some methods to deterministically provide bandwidth to the client sessions to mitigate the contention and starvation. Therefore, our storage design is essential for large concurrent multi-session workloads with intensive reads such as AI training.
2022 OCP Global Summit
With the advent of new application workloads related to Big Data including AI/ML, IoT, Video, security and many other machine generated data, there is a strong incentive for companies as well as governments to mine this treasure trove of data to extract value. Innovative companies taking on these storage challenges have tried storing Big Data into data lakes and moving them into locally attached storage for analysis but due to the sheer size of the data set this turns out to be too time consuming. Traditional network data storage also have been explored but overcoming inherent scaling and performance bottlenecks are difficult. DSS provides an innovative solution to the problem by designing purpose built storage that only targets these specific workloads.
Go back to Main Page