Hi. Welcome to Ingram Micro.

Please choose your role, so we can direct you to what you’re looking for.

If you’d like to learn more about Ingram Micro global initiatives and operations, visit ingrammicro.com.

Build enterprise data centers with cloud scalability using NVIDIA accelerated networking

November 02, 2021

Build enterprise data centers with cloud scalability using NVIDIA accelerated networking
The digital transformation trend impacts companies of all sizes everywhere—changing how organizations access, resource, process and protect their assets to meet evolving business and market demands. This trend has profoundly affected enterprise data centers too. No longer is a company’s data center a single repository in one geographical location. More often, it’s a combination of one or more data centers or a private cloud built on-premises, connected to one or more public clouds.
Adding to this complexity, the data created in today’s data center is growing exponentially, especially unstructured data—such as videos, images, audio, log files, or presentations—which experts predict will account for 80% of global data by 2025. As a result, companies face a growing challenge of managing all this information, particularly aggregating from silos and analyzing it across multiple data center locations.
In this NVIDIA blog series, we’ll explore the need for a paradigm shift in how data centers are architected, operated and scaled and show the critical role a modern data center plays in enabling digital transformation.
Hybrid cloud: combining the benefits of public cloud with on-prem
One of the primary appeals to cloud IT professionals is the ability to procure, configure and implement compute, storage and networking resources quickly, on an as-needed basis. Most data centers can’t compete with the cloud’s elasticity, mobility and availability, yet not all workloads within an enterprise are a fit for a public cloud environment. Some data and apps must remain inside the corporate firewall for data sovereignty, privacy or performance reasons.
In recent years, this data center transformation has evolved into the convergence of private and public clouds—on-premises enterprise clouds connected and integrated with public clouds.
In this hybrid cloud model, applications are virtualized into microservices and managed across data centers. Today’s hybrid cloud infrastructures leverage optimized hardware platforms, where GPUs power parallel, compute-intensive workloads, data processing units (DPUs) accelerate data movement and security and CPUs run general-purpose compute tasks, making the data center the new unit of computing.
Hyperconvergence delivers cloud flexibility to the enterprise
Hyperconvergence bridges the gap between the public cloud and private data center environments, and Hyperconverged Infrastructure (HCI) makes it easier to scale resources as required by business demands. At the same time, HCI provides data protection, VM mobility, high availability, rapid application deployment, data efficiency and cost savings. In an HCI environment, monolithic and proprietary systems are disaggregated and replaced with off-the-shelf server clusters, which can be easily scaled out. As a result, both infrastructure resources and business applications are virtualized. Additionally, all business applications, from virtual desktop interfaces, databases, big data analytics, even artificial intelligence, are provisioned and operated in this unified infrastructure, controlled by software with optimal efficiency and security.
Within the past several years, hyperconvergence came about as a ground-up rethinking of all the services that comprise the data center to deliver the agility and flexibility of the cloud as an on-premises solution. Hyperconvergence brings together existing storage, compute and network switching products into a pretested, prevalidated offering sold as a single solution. With a focus on the virtual machine (VM) or workload, all the elements of a hyperconverged infrastructure (HCI) support the VM as the basic construct of the data center.
Software-defined, hardware-accelerated networking
Networking is a crucial component of today’s data center infrastructure, providing high-performance, low-latency interconnects to unleash the maximum power of compute and storage systems. It must scale out easily and efficiently, facilitating data center expansion and cloud integration. It also needs to be open, software-defined to better support virtualization, security and automation. Only with a software-defined and hardware-accelerated network infrastructure can the data center achieve cloud-scale efficiency. Both hybrid cloud and HCI require expanded use of SDN and deploying those solutions at scale requires that the networking be software-defined and hardware accelerated, ideally accelerated by a SmartNIC or DPU.
NVIDIA meets these requirements with a complete stack of end-to-end data center solutions composed of NVIDIA ConnectX network adapter cards and NVIDIA BlueField DPUs in the servers, open switches in the fabric, cables and transceivers. NVIDIA networking enables the network infrastructure to be transparent to every workload through hardware-accelerated data transfer, virtualization and security and open networking. Additionally, it delivers predictable services with clear visibility and actionable insight. All these capabilities and services can be integrated into your data center infrastructure by APIs, libraries and SDKs—the very definition of software-defined.
Strong adoption and broad ecosystem
NVIDIA networking solutions are used in data centers and clouds in various industries. We’re able to supply channel partners with NVIDIA adapters, switches and cables either ordered stand-alone or as part of leading OEM servers. In addition, Ingram Micro Integration Services can build complete hardware systems that support a broad range of applications, including hybrid cloud, HCI and SDN. We’ll discuss these solutions and use cases in our upcoming blogs.
To learn more about these solutions, please visit our website at https://www.ingrammicro.com/.
To learn more about NVIDIA networking solutions, visit https://www.nvidia.com/en-us/networking and join NVIDIA GTC, the premier online conference for AI innovators, technologists and creators, Nov. 8–11, 2021.