Hammerspace, the innovator behind the global data environment, is redefining unstructured data architectures with the latest performance capabilities to further empower organizations to leverage any server, storage system, and network that best benefits their decentralized workflows.
“Throw whatever use case you want at it—whatever performance requirements. We can scale linearly to saturate any network, any storage, and any server type,” said Floyd Christofferson, VP of product marketing at Hammerspace. “Release 5 is the fifth generation of the software that is bringing to bear the best of what we have learned over the last few years: how to do this as a scale-out architecture.”
The Release 5 software iteration provides high-performance across data centers and the cloud, ensuring users and applications can depend on a globally shared, secured access network regardless of storage platform or location. Hammerspace negates the traditional notion of requiring data placement to be as local as possible to meet performance needs; instead, Hammerspace’s Parallel Global File System orchestrates data automatically and by policy in advance, making data present locally without having to accommodate for data placement time.
“The innovation at Mellanox was focused on increasing data center efficiency by providing the highest throughput and lowest latency possible in the data center and in the cloud to deliver data faster to applications and unlock system performance capability,” said Eyal Waldman, co-founder and previous CEO of Mellanox Technologies. “I see high-performance access to global data as the next step in innovation for high-performance environments. Hammerspace makes it possible to take cloud utilization and data utilization to the next level of decentralization of where data resides.”
Additionally, unstructured data workloads in the cloud can leverage as many compute cores as allocated, and as much bandwidth as needed—even going as far as saturating the network within the cloud for greater compute environment and application connection. Other features of the release include Backblaze, Zadara, and Wasabi support; scalability improvements; back-end performance enhancement; improved resiliency across large, distributed environments; and a new Hammerspace Management GUI.
Hammerspace is also introducing improvements to performance across interconnections within data centers, reducing friction between data ingestion sources (such as instruments and large compute clusters), and compute and storage environment optimizations to reduce data ingestion into storage time expenditures. Compatible with a variety of highly performant storage platforms, Hammerspace is able to saturate even the fastest storage and network infrastructures to maximize aggregate throughput and IOPS while simultaneously embracing ease-of-use. According to the company, this enhancement resulted in a 20% increase in metadata performance to accelerate file creation in primary storage use cases. It further offers accelerated collaboration on shared files and RDMA support for global data over NFS v4.2.
Targeted towards edge environments and DevOps workloads, Hammerspace’s high-performance server-local IO enhancements allow users to employ the local server to its maximum benefit. Hammerspace delivers 73.12 Gbits/sec performance from a single NVMe-based server, illustrating its performance capabilities in a similar comparison to direct-to-kernel access. The Hammerspace Parallel Global File System separates the metadata control plane from the data path, allowing the use of embedded parallel file system clients with NFS v4.2 in Linux, mitigating overhead.
For more information about Hammerspace’s latest performance capabilities, please visit https://hammerspace.com/.