The Diamanti platform consists of three key components: Diamanti Spektra, Diamanti Ultima and Diamanti D20 Series — all purpose-built for modern, distributed applications. In this second of the three-part blog series, we will discuss Diamanti Ultima, a patented technology that offloads storage and networking I/O traffic from the CPU, freeing compute and memory resources for modern applications. Diamanti Ultima accelerates storage and networking I/O while delivering traffic isolation and guaranteed quality of service (QoS).
What is Diamanti Ultima?
Diamanti Ultima is a pair of second generation PCIe based I/O acceleration cards that offload networking and storage traffic freeing up compute resources to power modern applications and deliver dramatically improved performance.
In a traditional server, the CPU is responsible for more than powering applications. It is typically also managing all of the I/O traffic for storage and networking. By offloading storage and network traffic, Diamanti Ultima drives 95% host utilization, making more processing power available for actual applications. This improves overall efficiency and reduces the server footprint, which in turn can provide 80% total cost of ownership (TCO) savings.
In addition to greater efficiency, Diamanti Ultima delivers 10x to 30x performance improvement with sub-100 microsecond latency and 1 million IOPS per node for I/O-intensive applications. The unique aspect of Diamanti Ultima is the policy-based quality of service (QoS) feature with which an application is assured a specific level of storage IOPS and network throughput at all times. This enables an application to operate at consistent performance and latency regardless of the number, type or size of other applications operating on the same node. These improvements are highly critical for modern applications that are highly distributed and scale-out on demand.
Integrated Storage, Networking and Security
Purpose-built for Diamanti Spektra, Diamanti Ultima integrates with Kubernetes via the standard Container Storage Interface (CSI) and Container Network Interface (CNI) plugins. Operating in tandem with ultra-fast all NVMe storage on the Diamanti D20 series, Diamanti Ultima also provides enterprise class storage services for disaster recovery (DR) and data protection (DP) of mission critical applications, including mirroring, snapshots, multi-cluster asynchronous replication, and backup and recovery.
Diamanti Ultima supports SR-IOV for high performance networking and implements container granular quality of service (QoS) policies to deliver guaranteed SLAs to applications. Diamanti Ultima can integrate with standard enterprise data center networking topologies, delivering both overlay and non-overlay network interfaces, static and dynamic IP assignment and support for multiple interfaces per pod.
With hardware-based queues, Diamanti Ultima isolates network and storage traffic at container granularity to provide maximum security and isolation for applications. Further, management and application traffic are separated to maintain control and data plane bifurcation. With support for volume encryption and self-encrypting drives (SED), enterprise customers can prevent data theft and protect against security breaches while not incurring significant performance penalties.
Get More Out of Your Infrastructure with Diamanti Ultima
With it’s patented I/O acceleration technology, Diamanti Ultima enables networking and storage traffic to be completely offloaded to dedicated PCIe processors, freeing up compute resources to deliver 80% TCO savings and deliver dramatically improved and consistent performance. Greater control of network and storage resource allocations across the cluster improves consistency, enables multi-tenancy and eliminates noisy neighbors.
To learn more about how Diamanti Ultima can supercharge your applications and Kubernetes deployment, watch the on-demand webinar discussing the latest and greatest features of Diamanti Spektra 2.4.
In the next blog post, we will discuss the Diamanti D20 Series, a set of validated modern hyperconverged platforms purpose-built to run stateful, stateless and AI/ML applications.