Supercharged Kubernetes: Part 3 – Diamanti D20 Series

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 third of the three-part blog series, we will discuss the Diamanti D20 Series.

Figure 1: The Diamanti Platform

 

Hyperconverged Infrastructure has transformed the way companies can deploy, scale, and manage the applications that power their businesses. It combines compute, networking, and storage with intelligent software all in one box, making it easy to manage the infrastructure and reduce the infrastructure footprint. Many applications such as databases, messaging, analytics, and Artificial Intelligence (AI) and Machine Learning (ML) are being modernized and deployed as containers in Kubernetes. The combination of containers and Kubernetes on hyperconverged infrastructure can provide organizations the simplicity, ease, and agility required to supercharge their modern applications.

When organizations deploy traditional hyperconverged infrastructure for containerized applications cracks begin to appear in its architecture. Production applications suffer from “noisy neighbor” problems which force users to over-provision the infrastructure, resulting in increased costs. And running containers inside virtual machines adds unnecessary layers of abstraction. Diamanti D20 series completely eliminates the noisy neighbor problem with Diamanti Ultima IO acceleration cards and delivers unprecedented application performance as a bare-metal Kubernetes platform. Our latest platforms truly help to shed the complexities of legacy infrastructure by providing significant performance improvements, increased security, TCO savings, and creates an easy adoption path for modern applications.

 

Diamanti D20 Series: Modern Hyperconverged Infrastructure

The Diamanti D20 series of modern hyperconverged platforms consists of D20, D20X, G20T, and G20P with varying configurations of Intel CPUs, memory, and NVMe storage. G20T and G20P also support Nvidia GPUs for AI/ML use cases. Powered by Diamanti Ultima acceleration cards, the Diamanti D20 series delivers the first and only Kubernetes solution integrated with a patented I/O-optimized architecture, delivering transformational application performance. Intesa Sanpaolo, a leading bank in Europe successfully embraced microservices and container architecture to migrate from monolithic to multi-tier applications with the D20 infrastructure.

 

Figure 2: Diamanti D20 Series

 

Diamanti D20X

Adding to the D20 family of modern hyperconverged infrastructure, the new D20X supports the latest 2nd Generation IntelⓇ XeonⓇ Scalable Processors (Cascade Lake). The recently released processors deliver an average of 36% higher computing power, with increased core counts, higher cache, and higher clock frequencies. This translates to more database transactions, improved application performance and faster deployments on a per rack unit (RU) basis, and investment protection in the long run. Diamanti D20X is packaged with Diamanti Ultima I/O acceleration cards and a wide range of NVMe drives and RAM options. Diamanti D20X is available with self-encrypting drives (SED) option ensuring unparalleled protection for critical applications.

 

Figure 3: Diamanti D20X

 

Diamanti G20

AI/ML is being broadly adopted across companies and is guiding decisions on everything from talent recruitment to bank loans. When looking at AI/ML from an economic perspective, companies need to ensure that they maximize ROI on expensive AI/ML infrastructure investment while supporting their business needs. To cater to these requirements, Diamanti launched the G20 series that enables faster time-to-insights and caters to the exponential growth in machine learning models generated by data scientists while keeping budgets under control. The G20 series has found increasing traction in the energy and services sector, helping these organizations to deliver new AI/ML-enabled business improvements.

Figure 4: Diamanti Machine Learning Platform

 

The G20 series combines the power of Diamanti Ultima, NVIDIA GPUs, and lightning-fast NVMe shared storage to provide a bare metal Kubernetes platform for data scientists to run ML workloads while maximizing resource utilization and return on investment. The G20 series consists of multiple configurations for model training and inference to cater to the varying AI/ML use cases. G20 is offered in two flavors:

  • G20T: With a fully-loaded configuration, the G20T is best suited for intensive AI/ML modeling as data scientists can utilize maximum compute capacity for training and retraining their models
  • G20P: G20P is intended for AI/ML inference to cater to the varying AI/ML use cases in production. G20P is available in three configurations: Small, Medium, and Large

The Diamanti Spektra management console enables data scientists to deploy and manage both GPU and CPU targeted workloads to the same cluster.

 

Figure 5: Diamanti Spektra Console

Diamanti D20 Series: Infrastructure For Modern Applications

Diamanti D20 Series with Diamanti Spektra and Diamanti Ultima together deliver purpose-built infrastructure for modern applications. This powerful combination works perfectly for stateful, stateless, and performance intensive workloads such as AI/ML, databases, messaging, and analytics. By eliminating unnecessary layers of abstraction, Diamanti empowers modern IT to scale quickly while delivering transformational application performance, reducing data center footprint, and total cost of ownership (TCO).

For more information about how Diamanti Spektra 2.4 can help your containerized applications thrive, please watch the on-demand webinar.