All articles

What did I Discover at HPE Discover 2024?


Kyle Davies

OCTO Round-Up

•  Jun 24, 2024

This year’s HPE Discover was the biggest yet, and HPE were the first company to host their keynote within the spectacular Las Vegas Sphere! 

There were over 15,000 delegates attending the event throughout the week, where, for the first time, HPE merged Discover with Aruba Atmosphere. 

The event took place in sunny Las Vegas, spanning 7 acres, 150+ demos and 100s of sessions, as well as great opportunities to have 1-2-1 sessions with the HPE leadership team(s) to dive further into their plans. 

As expected, the core underlying theme of the week was ‘AI’, but there were plenty of updates and announcements that took place within the Hybrid Cloud portfolio. For ease, I have split the remainder of this article into these two parts specifically. 

AI - Intelligence has no limits 

HPE has unveiled its latest innovations in artificial intelligence (AI) at HPE Discover 2024, showcasing how it is helping customers and partners accelerate their digital transformation. Antonio made a clear statement, “One day soon, you will be recording data on your balance sheet”, which is why HPE and the industry has gone wild over AI and how it can potentially unlock the next evolution of business. 

During the conference, Jensen from NVIDIA joined Antonio on stage, and it was great to hear them outline that AI is not new! We are just moving from instruction-based computing to intent-based computing. 

Some of the highlights from the event include: 

The solution is the first to combine NVIDIA's AI computing, networking and software with HPE's AI storage, compute and the HPE GreenLake cloud. It helps enterprises of any size to develop and deploy generative AI applications in an efficient, fast and flexible way. The solution uses the new OpsRamp AI copilot, which improves workload and IT efficiency for IT operations. HPE Private Cloud AI has a self-service cloud experience with full lifecycle management and comes in four configurations for different AI workloads and use cases.  

The core goal of this partnership is to combine the entire NVIDIA AI computing stack with HPE’s private cloud technology to equip customers and AI professionals with the most advanced computing infrastructure and services. This is to expand the frontier of AI, whilst lowering the risk and complexity. 

The solution: 

  • Allows use of proprietary data for inference, fine-tuning and RAG AI workloads 
  • Ensures data privacy, security, transparency and governance for enterprise needs 
  • Allows cloud-like experiences for ITOps and AIOps features to boost productivity 
  • Allows flexible consumption for future AI prospects and expansion using HPE Greenlake Cloud 

Upon first look, you may be thinking it is just HPE systems with NVIDIA accelerators, but there has been a lot more effort put in the background to simplify the complexity of AI workloads.  

NVIDIA AI Enterprise enables faster data science workflows and simplifies the creation and deployment of production-ready copilots and other GenAI applications. NVIDIA AI Enterprise comes with NVIDIA NIM, which provides user-friendly microservices for enhanced AI model inferencing, which support a seamless transition from proof-of-concept to secure deployment of AI models in various scenarios. 

HPE AI Essentials software, along with NVIDIA AI Enterprise and NVIDIA NIM, offers a set of preconfigured AI and data foundation tools with a unified control plane, which provide flexible solutions, continuous enterprise support and reliable AI services. For example; data and model compliance and expandable features ensure AI pipelines are compliant, explainable and reproducible throughout the AI lifecycle. 

This is shown by the image below: 

I managed to have a play with this on one of the demo booths, and within 5 clicks was able to deploy a private virtual chat bot. I know this isn’t the most complicated, but it shows the power of the automation efforts both NVIDIA and HPE have made in the platform to simplify the way we engage with the opensource models and ever-growing AI services catalogue. With another 5 clicks, I was then able to start fine-tuning my bot with a specific data set. In the past, getting IT Ops and your Dev/Data teams to collaborate on this could have been challenging, whereas now this is self-service for IT Ops to fulfil demand to the business.  

Underpinning the capabilities are T-Shirt-sized systems that arrive at your location, ‘plug-and-play', to start your desired AI initiative.  

To further optimise performance for the AI and data software stack, HPE Private Cloud AI delivers a fully integrated AI infrastructure stack, which includes NVIDIA Spectrum-X™ Ethernet networking, HPE GreenLake for File Storage and HPE ProLiant servers with support for NVIDIA L40S, NVIDIA H100 NVL Tensor Core GPUs and the NVIDIA GH200 NVL2 platform. 

One of the statistics used onstage was that using the HPE bundles above you could see up to a 5x cost efficiency for inferencing! 

As you can see above, this can dramatically reduce the time and effort needed to scope your AI inferencing, RAG and fine-tuning needs. Is it perfect? No… but it is a start and will aid in the accelerated playground and move to production for hybrid cloud AI services. 

  • Certified for NVIDIA DGX BasePod 

Another under the radar update was that HPE GreenLake for File Storage has achieved NVIDIA DGX BasePOD certification and NVIDIA OVX storage validation. This will provide customers with a proven enterprise file storage solution for accelerating AI, GenAI and GPU-intensive workloads.  

Moving forward, HPE have committed to be a time-to-market partner on upcoming NVIDIA reference architecture storage certification programs. 

OpsRamp's operations capabilities are integrated with HPE GreenLake cloud to provide observability and AIOps for all HPE products and services. OpsRamp now offers observability for the complete NVIDIA accelerated computing stack, including NVIDIA NIM and AI software, NVIDIA Tensor Core GPUs and AI clusters as well as NVIDIA Quantum InfiniBand and NVIDIA Spectrum Ethernet switches. IT administrators can use insights to detect anomalies and monitor their AI infrastructure and workloads across hybrid, multi-cloud environments. The new OpsRamp copilot leverages NVIDIA’s accelerated computing platform to process large datasets for insights with a NLP assistant, improving efficiency for operations management. OpsRamp also integrates with CrowdStrike APIs so customers can view a unified service map of endpoint security across their entire infrastructure and applications. 

Just imagine the complexity of your AI service stack, data stack, compute, storage and networking. Finding the root cause of an underperforming service could become challenging therefore this OpsRamp copilot will narrow down your field of view to enable your teams to respond more effectively. 

To summarise, OpsRamp now has AI observability, AIOps Copilot and deep integration with CrowdStrike.  

Hybrid Cloud – Unlock Capabilities 

HPE has also announced its next phase of capabilities for their hybrid cloud portfolio at HPE Discover 2024, demonstrating how it is enabling customers and partners to harness the power across any environment, from the edge to the cloud. Some of the key announcements are below. 

One statistic that stuck with me was that over 400,000 HPE systems out in the wild are running legacy services or traditional virtualisation workloads. These workloads are only providing 7% of value back to the business. HPE outlines that when modernising to the new generation of systems you can replace 8 legacy systems with one system, thus reducing your operational overheads, power requirements etc. When you take this into context, alongside their partnerships with VMware, Nutanix etc the announcements below start to give you a wider view of their strategy to provide flexibility and choice, whilst having concise management capabilities. 

  • HPE Virtualisation Capability 

HPE announced the expansion of its hybrid cloud offerings, by adding the HPE virtualisation capability for HPE Private Cloud. It combines open-source kernel-based virtual machine (KVM) with HPE’s cluster orchestration software to support the high performance and availability required for demanding enterprise workloads.  

Their virtualisation clusters are managed through a cloud-based control plane (GreenLake), which removes the need to dedicate valuable resources to on-prem management, whilst still being designed to remain highly available, even when disconnected from the cloud.  This enables automated Day 0 and Day 1+ operations, two-click VM provisioning with integrated data protection, and automated lifecycle management.   

One demo I got to see in action was a 3-click migration process for virtual machines to move onto the HPE Virtualisation platform. HPE have put in a lot of time into the orchestration layer, in order to simplify the potential migrations. I managed to move instances from various hypervisors and public clouds using this same 3 click approach. 

HPE Private Cloud enables virtualised workloads, cloud-native workloads, and AI workloads; supporting virtual machines, containers, and bare metal. The virtualisation addition now provides HPE with an end-to-end framework for their hybrid cloud story, which they announced many years back.  

Why are HPE investing so heavily in this? Well, they believe that two thirds of workloads are not public cloud-ready. Therefore their offering caters to the wider audience whilst providing flexibility and choice. Basically, make cloud come to you! 

  • Complete Stack 

With HPE’s addition of its new virtualisation technology, HPE can offer a full suite of enterprise-grade capabilities and services across the hybrid cloud technology stack, including: 

  • Flexible private cloud infrastructure with AI-driven automation Network virtualization and segmentation (HPE Aruba Networking) 
  • Data portability with consistent experience across private & public cloud, using SDS with HPE Alletra MP 
  • Full range of runtimes (virtual machines, containers, bare metal) and AI frameworks 
  • Multi-cloud, multi-vendor cloud services for IT operations management with OpsRamp 
  • Data protection and cyber resiliency with Zerto 
  • Sustainability reporting with Sustainability Insights Center 
  • Integrated, platform-based cloud management layer powered by HPE GreenLake 

Overall, the event was insightful, and I look forward to obtaining pricing and further details on the above over the coming weeks.  

If you would like to learn more about CDW and our HPE partnership, feel free to reach out to your aligned account manager, or contact us to discuss your requirements.

Subscribe to email updates

Related insights

Dell Technologies World 2024 AI Edition Summary
  • Hybrid Platforms

Dell Technologies World 2024: AI Edition

Rob Sims analyses Dell’s new offerings from the Dell Technologies World 2024 event, and addresses why AI is a central part of Dell’s future strategy

Read article
  • OCTO Round-Up

Office of the CTO Round-Up: VMware Explore 2023, Las Vegas

VMware took to the big stage over in Las Vegas, Nevada at their Explore event. Read about the major announcements, initial thoughts, and follow up with more detail between now and November's event.

Read article
  • OCTO Round-Up

Red Bull Event July 2023

Oracle Red Bull invited Modern Workspace CDW Chief Technologist, Tim Russell, to their MK7 location for an opportunity to talk about Digitalisation at the Red Bull Innovation Day 2023.

Read article