Historically, Artificial Intelligence (AI) platforms were used primarily by progressive companies with the ability to invest extensively in complex systems to garner an advantage in their respective markets. The financial services and telecommunications industries who consume and analyze vast amounts of data were early adopters in AI-Machine Learning (ML) primarily because of their previous development of data networks and implementation of massive-scale computing. Today decision makers in those industries continue to seek out solutions that can help them grow their businesses and reduce costs.
“This change in how business operates will be key for the years ahead, especially as the marketplace remains in flux and customer demands for contextually relevant experiences increase. Artificial intelligence technology may not be mainstream yet, but it’s only a matter of time.”Bill Porto, May 10, 2018
If AI-ML was solely the domain of specialized data scientists, unattainable by most IT organizations it was at least partially because these AI-ML workloads were run-on purpose-built infrastructure which was extremely complex, expensive, and very difficult to design, optimize, and support. Highly specialized teams of engineers were required to create a functional environment where data scientists, who generally had little or no knowledge of the underlying platforms running the actual AI-ML workloads. However, recently there has been a major advancement from industry leaders like HPE, VMware, and NVIDIA providing management tools, software, and hardware that eases the user experience and which allow enterprises who wish to leverage the advantages of AI-ML to deploy such functionality easily. Now customers can run these AI-ML workloads in the core infrastructure with centrally managed IT instead of sprawling ‘peripheral’ IT deployments.
On March 9, NVIDIA and VMware announced the AI-Ready Enterprise Platform that includes NVIDIA AI Enterprise suite optimized for VMware vSphere, the data center standard for virtualization, and their OEM partner ecosystem. This is a game changer that democratizes AI for enterprises and brings it to core mission critical workloads. The combination of VMware vSphere, NVIDIA GPUs, NVIDIA AI Enterprise suite, and HPE ProLiant DL380 rack optimized servers defines the new industry standard AI servers. HPE has partnered for decades with VMware and NVIDIA in delivering industry leading integrated solutions for a variety of use cases. In the next section of the blog, we will explore how each one of these industry leaders contribute to bringing AI to enterprise workloads.
Let’s start with the top reasons why customers choose the HPE ProLiant DL380 to be their intelligent AI-ML compute foundation for addressing key IT challenges.
- Configure and optimize workloads at the click of a button
Enabled by HPE Workload Matching, you can quickly and efficiently choose from a selection of preconfigured workload profiles. With a click of a button, these profiles allow you to match the characteristics of your workload to a subset of BIOS tuning options that we’ve determined are best suited to maximize performance.
- Monitor and optimize workload performance in real time
HPE’s Workload Performance Advisor monitors key server performance metrics and provides users with real-time tuning advice for select BIOS settings. Tuning advice is based on actual server resources that are being used while running your workload.
- Secure systems, even when on the move
HPE Server Configuration Lock protects systems in transit and when deployed in insecure locations such as retail stores, outdoor access points, IoT platforms, and more. You can rest assured that the system has not been tampered with and does not get deployed in a compromised state. This way, systems shipped from one location to another are not vulnerable to attack in the supply chain.
- Seamlessly monitor and manage servers
HPE iLO 5 management provides detailed insights into server status and operations, allowing system administrators to monitor and manage servers remotely anywhere, anytime. Utilizing a wizard-based embedded server provisioning utility, iLO 5 provides intelligent provisioning to make it easy to take your server from out-of-the-box into production.
- Use AI to maintain the ideal operating environment
HPE InfoSight for Servers is a unique feature for HPE ProLiant servers and other HPE infrastructure that takes a comprehensive approach to data collection and analysis, collecting, and analysing millions of sensors per second. Using AI operations, InfoSight continuously learns from the telemetry and develops an understanding of the ideal operating environment for every system, workload, and application. InfoSight uses its predictive analytics and recommendations engine to analyse health, configuration, and performance data to predict and prevent problems and improve performance.
- Industry’s most comprehensive and complete GPU support
HPE has one of the most extensive portfolios of GPU support. The newest addition NVIDIA Ampere 100 is supported across ProLiant, Synergy, Apollo, and Edgeline servers.
- Key Integrations between OneView and VMware vCenter providing superior user experience
HPE has one of the largest vSphere install bases and has several mature software plug-ins between HPE OneView and VMware vCenter, vRealize, and SDDC Manager. The integrated life cycle management allows customers to update firmware for both hardware and software from a single user interface making this a fully streamlined process and provides superior governance over version drift.
VMware and NVIDIA continue to improve the effective use of AI/ML technology across several key vertical segments. Businesses and even entire industries seek to harness the power of AI to extract new insights from massive data sets, both on-premises and in the cloud.
The NVIDIA Ampere architecture, designed for the age of elastic computing, delivers the next giant leap by providing unmatched acceleration at every scale, enabling these innovators to do their life’s work. NVIDIA AI Enterprise is an end-to-end, cloud-native suite of AI and data analytics software, optimized, certified, and supported by NVIDIA to run on VMware vSphere with NVIDIA-Certified Systems. It includes key enabling technologies from NVIDIA for rapid deployment, management, and scaling of AI workloads in the modern hybrid cloud.
VMware has taken virtualization to great heights that it has become the foundation of any cloud operating model including the major public cloud vendors. Enough has been written about the efficiencies that VMware brings by virtualizing compute, storage, and networking across data centers. VMware launched vSphere 7.0 U2 recently that included support for virtualizing GPUs. In this version support for NVIDIA A100 GPUs is included along with capability of Multi Instance GPUs (MIG). A VM can now attach a MIG-backed vGPU with a dedicated hardware allocation of streaming multiprocessors (SMs), framebuffer (on-GPU) memory, and the various hardware paths to them. Learn more about the technology details.
Customers and partners can now leverage this one-stop-shop to accelerate their AI journey. AI has proven to help organizations deliver better business outcomes, improve productivity, and is an integral part of data center modernization. The leader in GPU and AI solutions (NVIDIA), leader in virtualization and hybrid cloud (VMware) and industry’s most reliable, secure, and versatile server in HPE ProLiant DL380 are coming together to help our mutual customers in this journey. Get started today!
Michael Porter, Senior Product Manager, HPE
Srikanth Vijayaraghavan, Global Technical Alliances, VMware