How NVIDIA AI Enterprise Suite Simplifies AI Deployment

This week GPU market leader and AI thought leader NVIDIA revealed its NVIDIA AI Enterprise, a comprehensive suite of enterprise-grade AI tools and frameworks. The software brings together everything a business would need to deploy an AI system and is optimized, tuned and certified by NVIDIA, exclusively with VMware vSphere 7 update 2, also announced this week. By running on a de facto standard platform, like vSphere, NVIDIA is making it easy for organizations to adopt AI Enterprise and then scale it as needed. NVIDIA truly simplifies AI deployment here.
AI Enterprise is designed to run on NVIDIA-certified systems based on its A100 Tensor Core GPUs. As is the case with most NVIDIA systems, the company has partnered with most of the mainstream server manufacturers and customers have a choice of purchasing hardware from Dell Technologies, Hewlett-Packard Enterprise, Gigabyte, Inspur and Supermicro. While the innovation is in the integrated software, the hardware platform certainly does matter.
AI Enterprise is optimized for vSphere
VMware has added support for NVIDIA GPUs in its latest release of vSphere and gives IT pros the ability to use vSphere Lifecycle Manager to see the virtual images. VMware’s Tanzu can be used to support Kubernetes, which is important today as an increasing number of AI workloads are deployed in containers. The latest release of Tanzu supports the VMware NSX Advanced Load Balancer to enable layer four load balancing of Kubernetes clusters. Also, vSphere 7 Update 2 now supports Confidential Containers for improved security.
AI is viewed by many, I included, to be the most transformative technology since the birth of the Internet. AI is rapidly being infused into everything and is changing the way we work and live. Cars are fast-tracking to self-driving, speech analytics are able to take what we say and turn it into actions, contact centers are able to service customers faster and doctors can find the smallest anomalies in MRIs saving thousands of lives. This is why almost every business and IT leader I interview from enterprise-class companies have AI on the near-term roadmap.
Customers now have options other than bare metal for AI
To date, AI systems have run on bare-metal servers, often purchased departmentally. This can be a very expensive and time-consuming way of deploying the technology as the systems are dedicated to one purpose. Also, scaling is limited to the hardware platform on which it’s running. Running on vSphere makes it easier to share resources and scale out when performance requires. As AI shifts from being a line of business owned to a resource that IT manages, having the system run vSphere provides the same level of agility and dynamism found in other IT systems.
The new AI Enterprise solution from NVIDIA is essentially an engineered system that removes much of the complexity of having to deploy all of the components independently and then tweak and tune things. This process can often take months and still lead to failures from having to manually provision and manage the disparate systems. Also, where there is a choice of software, customers need to run different tests and scenarios to ensure compatibility. With AI Enterprise, NVIDIA has taken all of the heavy lifting out of the deployment process, enabling organizations to start their AI initiatives months faster.
NVIDIA is a de facto systems company
I referred to NVIDIA in my opening paragraph as a GPU manufacturer, but in some ways, that statement is incorrect. NVIDIA is a systems company that uses GPUs as its building blocks. Companies can’t buy a GPU and use it right away. It needs specific hardware and software to make it work. It’s like a CPU; one can’t just buy a new CPU and start working. There needs to be a computer, operating system and applications to make it useful. NVIDIA does this with accelerated computing systems.
These engineered systems are why NVIDIA has become the market leader in GPUs. It has products such as Drive, used in self-driving cars, Metropolis for video analytics, Maxine for collaboration AI, Clara for medical imaging and DGX used in machine learning. These are just a few of the engineered systems that NVIDIA has put together to enable customers to maximize their investment in its technology.
As AI initiatives move out of the lab and into the mainstream, AI Enterprise will play a key role in helping enterprises across all industries accelerate their use of advanced technology to make better, faster decisions.
Zeus Kerravala is an eWEEK regular contributor and the founder and principal analyst with ZK Research. He spent 10 years at Yankee Group and prior to that held a number of corporate IT positions. Kerravala is considered one of the top 10 IT analysts in the world by Apollo Research, which evaluated 3,960 technology analysts and their individual press coverage metrics. 
The post How NVIDIA AI Enterprise Suite Simplifies AI Deployment appeared first on eWEEK.