QLogic Inc. is collaborating with Cavium Inc. to develop a joint solution aimed at accelerating performance enterprise storage, cloud storage, and Big Data storage applications. The solution is a combination of QLogic’s low latency 25Gb Ethernet (25GbE) RDMA-enabled NICs (RNICs) and Cavium’s high-performance 48 core ThunderX processors. Not only will customers see accelerate performance in bid data application such as Ceph and Hadoop, they will also have more flexibility and a better total cost of ownership.
QLogic Inc. is collaborating with Cavium Inc. to develop a joint solution aimed at accelerating performance enterprise storage, cloud storage, and Big Data storage applications. The solution is a combination of QLogic’s low latency 25Gb Ethernet (25GbE) RDMA-enabled NICs (RNICs) and Cavium’s high-performance 48 core ThunderX processors. Not only will customers see accelerate performance in bid data application such as Ceph and Hadoop, they will also have more flexibility and a better total cost of ownership.
Both Cloud and Big Data rely heavily on direct attached storage platforms. The problem with this is that it doesn’t allow the compute or storage to scale independent of one another. The new joint solution will address this scalability issue by providing a “disaggregated storage architecture that brings modularity, improved efficiency, and higher performance to big data and cloud storage applications.”
QLogic 25GbE RNICs deliver heterogeneous remote direct memory access over Converged Ethernet (RoCE), RoCEv2 and iWARP. And Cavium’s thunder family is a 64-bit ARMv8-A server processor designed for next-generation data center and cloud applications. QLogic’s 25GbE RNICs offload ThunderX platform workloads to deliver a cut-through, scalable and programmable architecture. This architecture accelerates object, block and file I/O, which will accelerate massively distributed scale-out storage applications.
Benefits of the joint solution include:
- Up to 2X system performance improvement compared to scale-out DAS based architectures
- Ability to independently scale compute and storage resources to petabyte-class databases and beyond
- Significantly reduced latency for distributed file systems accessing data over an RDMA-enabled fabric—without burdening the compute infrastructure
