Introducing Seagate ClusterStor Small Block Accelerator: Accelerate HPC Applications on Intel Lustre and IBM GPFS
SessionStorage
Session ChairJackie Kern
Presenter
Event Type
Exhibitor Forum
Accelerators
Clouds and Distributed Computing
Effective Application of HPC
State of the Practice
Storage
TimeThursday, November 17th11:30am - 12pm
Location155-F
DescriptionThe common workload in High Performance Computing (HPC) consists of mixed block sizes from small to large. Parallel file systems such as Lustre and GPFS deliver high throughput with large block data but can become disk seek-limited with small block data.
In this talk, we will discuss the Seagate’s new ClusterStor Small Block Accelerator (SBA) that uses our advanced caching technology to enhance performance for those mixed applications workloads.
We will discuss technology and share results.
The SBA/NytroXD solution suite is a collection of storage acceleration tools that combine spinning and solid state disk storage media with caching technologies to intelligently accelerate block and file-based data access. It is part of the Seagate ClusterStor Architecture that extends the functionality of the storage system’s cache with SSDs and more persistent data with spinning media. This yields acceleration in the context of small blocks and also increases the throughput of the application.
In this talk, we will discuss the Seagate’s new ClusterStor Small Block Accelerator (SBA) that uses our advanced caching technology to enhance performance for those mixed applications workloads.
We will discuss technology and share results.
The SBA/NytroXD solution suite is a collection of storage acceleration tools that combine spinning and solid state disk storage media with caching technologies to intelligently accelerate block and file-based data access. It is part of the Seagate ClusterStor Architecture that extends the functionality of the storage system’s cache with SSDs and more persistent data with spinning media. This yields acceleration in the context of small blocks and also increases the throughput of the application.
