The reasoning behind HPE’s acquisition of Juniper could either be a mundane, economy-of-scale strategy or an attempt to take over Juniper’s AI networking technology. Alternatively, there might be a more ambitious plan at play.
Hardly any of the corporations I have interacted with can respond to the query “Why would HPE purchase Juniper?” The same goes for the competitors of either firm, Wall Street stakeholders, and even employees of the two corporations.
Any attempt to justify the transaction usually unveils three distinct interpretations of an old saying. It may imply that “the total is more than the sum of its parts,” as the proverb goes. Alternatively, it suggests that “the total can be operated at a lower cost than the sum of its components,” which is usually Wall Street’s viewpoint. It may also imply “the total requires an additional component,” the media’s perspective. The choice we make determines the fate of both companies.
The second perspective is the most common. Ideally, a successful M&A is predicated on the redundancy of staff when two companies merge, resulting in significant cost savings from layoffs. There’s no radical shift in strategy or product plans. It’s all about economies of scale. Albeit, such consolidation is usually a reaction to commoditization which means, at best, the companies are just managing a better position while spiralling downwards.
The rationale behind the collaboration between HPE and Juniper is asserted in all three alternative suggestions for understanding the ‘why’ of the business deal. Surprisingly, all these alternatives are grounded on a principle of symbiosis rather than simple functional efficacy. The proposition is that Juniper’s offerings can augment HPE’s value, and likewise, that of Juniper can be enhanced by HPE offerings. Or, the confluence of the offerings from both can deliver better results for everybody. All these alternatives would hinge on the ‘offering’ that is being incorporated. The real question, then, is what this offering is, to which there are two potential answers.
The first one is basic – HPE craves the AI technologies that Juniper owns. This seems a bit far-fetched at the outset, as Juniper’s offerings go well beyond AI. Is buying up a hardware store the solution when you need just a nut or bolt? Nevertheless, delving deeper into this point lends it a different perspective. The persistent issue of operational complexity returns here. Both data center and network operations grapple with numerous challenges, such as the risk of human errors, security concerns, high costs, inertia, among others. Juniper’s AI solutions, encompassing everything from Mist AI to Marvis Virtual Network Assistant, have been given the thumbs up by businesses as a useful tool for streamlining operations. Another key Juniper offering, Apstra, is all about conceptualizing a virtual data center where AI extends to the data center. This could allow HPE to build integrations between their offerings and those from Juniper. Could this be what HPE is after?
Possibly, but there is still one pressing question – why didn’t HPE directly acquire the AI they needed? This brings up the second potential answer: HPE would’ve needed a reason to acquire the networking equipment as well. One possibility could be HPE’s belief in their ability to extract more value from Juniper gear than Juniper can. A conviction shared by enterprises seems to support HPE’s case. In the final quarter of 2023, nearly 90% of enterprises reported that their network planning and procurement decisions were primarily guided by the vendor of their data center. The realms of application software, platform software, middleware and computing are more impactful than network-specific issues. HPE might well be in an advantageous position to sell Juniper’s gear effectively….provided they understand this factor and integrate it into their marketing initiatives.
But one challenge staring HPE in the face is this: despite their greater strategic control as IT products vendors, all elements of data centers are rapidly becoming commoditized much like networks. Evidently, influence doesn’t necessarily equate to building a business case. In essence, HPE needs to breathe some life into certain stagnant areas of IT. They need to develop a set of new business cases which will set the stage for the launch of new applications, in turn laying the groundwork for the onset of new computing norms and more networking. Following this value chain might prove challenging, and there is no clear indicator at the moment that either HPE or Juniper will be to pull it off without a hitch. However, certain signs about what a business case needs to look like could provide hints about what HPE is envisioning.
Over the past seven decades, there have been three instances where corporate expenditure on IT has surpassed GDP growth. These occasions coincide with the mainframe, minicomputer, and PC eras, respectively, with each stage bringing IT closer to the individual user. However, this progress has halted and the question begs, what will reignite it? The solution may lie in empowering activities at their point of execution, facilitated by real-world, real-time computing.
Consider a large-scale digital twin of our real world, composed of several smaller digital twins such as rooms, buildings, cities and towns – all interconnected. Consumers and employees interact with both the physical and digital realities, and their digital presence enables the delivery of the exact information they need, at the right place and time. What value could this bring in terms of productivity and quality of life? The complexity of IT and network operations has increased, shifting from punch card data entry to real-time input with computer-to-computer communication within organizations. Imagine the complexities involved in initiating a system where the real world and its digital counterpart are intertwined? The risk of global failure on a lightning-fast scale is unacceptable. We require a superintelligence to back us up.
This is the level of support that AI offers, as signaled by Juniper’s recent announcement of its AI-Native Networking. Juniper has also forayed into the computing arena with their universal data center virtualization tool, Apstra, which is now incorporated in its AI-Native Networking solution. AI-Native Networking may not be vital to present network/IT operations enough to make HPE’s acquisition an immediate success, but it could very well be if HPE emphasizes point-of-activity empowerment. If so, HPE and Juniper’s combination could play a pivotal role in bringing this vision to life. It’s possible that Cisco’s recent partnership with Nvidia to incorporate AI into their networks is driven by apprehension over the potential of AI in this context.
The obvious concern with the proposed HPE/Juniper merger is the possibility of it taking a year to finalize. This provides competitors ample time to discern the potential impact of the merger and devise strategies to counter it. Nevertheless, HPE and Juniper have the same time frame to explore innovative solutions and envisage scenarios that competitors might find difficult to counter.
The danger inherent in this situation is the inability to actualize dreams that are non-existent. At present, we are unaware of HPE’s visionary objectives. The mere fact that most reasons given for acquiring Juniper aren’t compelling doesn’t rule out the possibility that HPE might have selected one of those reasons. The opportunity within the present real-world context has not yet become evident this year—even though IBM’s threat to HPE has. Neither HPE nor Juniper noticed this real-time opportunity previously. Is it possible for them to transition from tactical maneuvers and quarterly accounts and pursue their dreams? A quote from Robert Kennedy is apt here: “Some men see things as they are and ask ‘Why?’ I dream of things that never were and ask ‘Why not?'” To restate it within this context: HPE, why not?
