HPE Doubles Down on HPE Nonstop Compute

Twelve years ago, Hewlett Packard Enterprise (HPE), formerly known as Hewlett-Packard Company, announced a strategic initiative to extend the HPE Nonstop solution to the Intel® x86 architecture. Subsequently, two years later, HPE introduced the HPE Nonstop NS7 X1, making the HPE Nonstop on x86 a reality while redefining continuous availability and scalability for x86.

Today, HPE is announcing the fifth generation of bare metal systems based upon the x86 with the introduction of the HPE Nonstop Compute NS5 X5 and the HPE Nonstop Compute NS9 X5. These are the latest products in the entry class and high-end categories respectively.

As part of these two new solutions, HPE Nonstop is doubling down in areas that matter to our customers. On the high-end system, the maximum memory capacity doubles from 4 TB to 8 TB of RAM. On the entry class solution, the maximum memory capacity is also benefiting with a 2X increase in memory capacity, moving from 256 GB to 512 GB of RAM. This increase will enable customers to process more transactions and grow their business.

In the networking domain, we are significantly boosting the maximum raw I/O connection speeds across the board. Notably, entry-level system customers will experience a tenfold increase in maximum raw Ethernet speeds. While in other places there will be 2X to 2.5X increases.

To help dig a little deeper and uncover the finer details, Eric Brown is joining the conversation with his thoughtful and insightful questions. Eric brings an impressive 30 years of experience working with HPE Nonstop systems, where he has designed and architected innovative solutions across a range of industries—from the demanding Oil and Gas sector to the fast-evolving Payments and Financial services. What really makes Eric stand out is his natural curiosity and his exceptional ability to challenge the status quo. He’s always thinking two steps ahead, asking the tough questions that push the conversation forward. It’s these very qualities that make him such an invaluable voice as we take a closer look at the new HPE Nonstop Compute solutions.

I’ve heard many references to Starship. What is Starship and how does it relate to the NS5 and NS9 you introduced here?

That’s a great starting point. Starship is the internal code name of the program that created the NS5 X5 and NS9 X5 systems. It encompasses all the functional elements used to create the two new products. Going forward now, we will use the newly introduced names of the HPE Nonstop Compute NS5 X5 and the HPE Nonstop Compute NS9 X5.

Why Starship? I mean the NS5 and NS9, and what were some of the driving factors behind it?

For customers who have been around HPE Nonstop for a while, they would have seen the shift of these systems using proprietary hardware to industry standard hardware. As such, we track the industry movement. The NS5 X5 and NS9 X5 are the next generation.

Mark, what are the most significant new features and innovations in the HPE Nonstop system being announced in June 2025? How do these features differentiate this system from previous versions?

That’s a big question. Let me break it up into several pieces starting with the system interconnect. At the heart of the HPE Nonstop systems is the system interconnect. This is the part of the system that connects the compute, storage, and networking together allowing them to operate as a single system. For the NS9 X5 we use InfiniBand High Data Rate (HDR) technology. We’ve had significant advancements in our system interconnect since the introduction of the NS7 X1. Take a look at Graph 1 to see how the interconnect speeds have progressed.

Next we have the compute microprocessors. The NS9 X5 CPUs will be based on HPE ProLiant DL380s with Intel® Xeon Gold 6400 series microprocessor. For the NS9 X5, we complement the Intel Xeon microprocessor with either 256 GB or 512 GB of system memory. While it is doubling the maximum capacity of its most recent predecessor, the NS8, it also represents a more than 2.5X increase since the first two generations of the x86 based systems. Graph 2 shows how memory options have evolved over the five generations of HPE Nonstop x86.

So, Mark from an integration and compatibility perspective, how do the new HPE Nonstop systems integrate with existing IT infrastructure?

Thanks for the insightful question, Eric – let’s dive into that. The new systems are designed for efficient integration. We’ve introduced several enhancements to ensure compatibility with modern data centers, supporting a range of industry-standard protocols and configuration options. This enables more straightforward deployment alongside your existing infrastructure, minimizing disruption and optimizing performance.

On the entry class systems, the network interconnect speed has increased from a single offering of 1 Gbps copper interface to a choice of two Ethernet 10G connection types: a 10G copper and a 10G MMF fiber. Both have been specifically designed to work exclusively with the NS5 X5 system.

For the high-end systems we maintain the 10G copper and 10G MMF fiber interfaces while at the same time introducing a new 25G MMF fiber solution. This 25G connection should play very nicely with our customer’s modern data centers.

There are additional improvements as to how our HPE Nonstop systems integrate into existing IT infrastructure. Specifically in the maximum raw I/O bandwidth of the Fibre Channel interfaces. On both platforms, we have doubled the speed from 16 Gbps to 32 Gbps. This facilitates how the HPE XP Enterprise SAN storage and back-up storage solutions connect to the HPE Nonstop systems.

Graph 3 provides a concise overview of the enhancements in Fibre Channel and Ethernet interconnects available on the entry class and high-end systems.

What OS do we need to run Starship?

We have a new release of the HPE Nonstop L-Series OS software L25.02. This is the base release that the NS5 X5 and NS9 X5 require.

Are there any compatibility considerations that organizations should be aware of when upgrading to this new system?

HPE Nonstop continues to provide our customers with x86 instruction set compatibility. This provides 100% binary compatibility while using the same compilers. Customers using native x86 compilers can move their binaries and run them on the new platforms. Or they may choose to recompile to take advantage of the latest optimization offered by the x86 compilers.

For those running code 100 objects and using the original HPE Nonstop CISC compilers, we maintain the ability to run those applications in either accelerated or interpreted modes.

For either approach, customers can confidently continue using their existing software while taking full advantage of the latest technology, enjoying a smooth and seamless experience.

Mark, what can you tell me about performance?

The NS9 X5 provides 15% more performance capacity than the NS8 X4 for similarly configured systems. This is based on HPE Nonstop internal benchmarking.

Will customers be able to incorporate an NS9 into my existing NS7 or NS8 clustering solution?

Yes indeed. With the introduction of the NS9 X5, HPE provides native InfiniBand Cluster with the previous two generations of HPE Nonstop systems.

And finally Mark, I gotta ask you — did you cut corners?

Maintaining the HPE Nonstop fundamentals of availability, scalability and reliability is core to what we do.

Eric, thank you for taking time out of your busy schedule to help The Connection readers get a firsthand perspective of our two new systems, the HPE Nonstop Compute NS5 X5 and the HPE Nonstop Compute NS9 X5. Customers can find out more about these new systems at https://www.hpe.com/info/nonstop