Benchmark From Nikhef, Surf, and Fungible for Performance Storage Solution Using FS1600 Node

Fungible Inc. announced they have achieved the highest performance between a single server reading data from a single storage target, clocking in at 6.55 million IO/s.

Also utilizing technology from Juniper Networks, Inc., the result was based on tests from a single 64-core AMD server connecting to a single company’s FS1600 storage node, marking the FS1600 as a fastest storage node.

The all-flash NVMe-oF storage node achieved nearly 2x better performance than the nearest competitor. The collaboration among Fungible, Nikhef, a partnership between the Institutes Organisation of the Dutch Research Council and 6 universities, and SURF, the collaborative ICT association of Dutch educational and research institutions, is designed to explore high-performance storage for physics and related HPC science.

“We are excited about the unparalleled, near-native performance we were able to achieve by offloading as much as possible. These results show great promise for the FS1600 and client-side DPUs,” said Jouke Roorda, engineer, high performing systems, Nikhef.

“The Fungible FS1600 will be recognized as the start of a new era in storage, thanks to its unique DPU capabilities. We see a bright future for these devices and DPUs in general,” said Tristan Suerink, IT architect, Nikhef.

“SURF accelerates Dutch research by investing in ICT innovation and we are eager to see how researchers will benefit from Fungible’s high performance storage solutions,” said Raymond Oonk, Sr. advisor, SURF.

“We are proud to be a part of this performance milestone. The results achieved are truly ground-breaking and will have far-reaching implications for data centers across the globe,” said Raj Yavatkar, CTO, Juniper. “As enterprises modernize data center infrastructures to accommodate innovative new business models for the digital economy, it is critical that we have next-gen storage solutions that can keep up with future market demands.“

As a result of this performance, data-centric workloads can be consolidated, leading to an increase in utilization of storage media and cost per IO/s decrease of at least 2x compared to existing SDS solutions.

“What we are achieving in the lab with Nikhef and SURF can be deployed throughout the world,” said Pradeep Sindhu, CEO and co-founder, Fungible. “We believe customers, partners and research institutions can push innovation boundaries with the Fungible Storage Cluster to reduce processing time and advance experimental activities in accelerator-based particle physics and astroparticle physics. Ultimately, it is revolutionizing the performance, economics, reliability and security of scale-out data centers.“

Powered by the company’s Data Processing Unit (DPU), the partnership leverages the Fungible Storage Cluster, which is a high performance, secure, scale-out, disaggregated storage platform, scaling linearly to 300 million IO/s in a 40RU rack, and extending further to many racks.

Each FS1600 storage target node is powered by two F1 DPUs and packs 24 standard NVMe SSDs delivering an aggregate of 15 million IO/s in a 2RU form factor. The FS1600 supports NVMe over TCP and enables enterprise data services such as data durability, data reduction, data security in-line and at line-rate, all selectable on a per volume basis.