HPE acquires Juniper: surprising, but makes quite a bit of sense

HPE acquires Juniper: surprising, but makes quite a bit of sense

The news that Hewlett Packard Enterprise (HPE) acquires Juniper Networks is certainly surprising. After all, with Aruba, HPE already has a business unit focused on networking. However, Aruba and Juniper are different in many ways, making this proposed acquisition not all that strange from HPE’s perspective. Whether employees and customers of Aruba and Juniper will be very happy about it, of course, is another story. We dive a little deeper into this matter in this article.

When we saw the push notification of HPE’s likely acquisition of Juniper Networks come in, we were mostly surprised. First, because this is an acquisition that we didn’t see coming right away. Aruba and Juniper are the two highest-ranking networking vendors in Gartner’s Magic Quadrant. A large player acquiring a somewhat smaller one is fairly common, but two large ones merging (albeit under the HPE umbrella) is still pretty special.

Second, because while $14 billion is a lot of money, it also indicates that hardware is worth much less than software, relatively speaking. Compare it to Cisco’s acquisition of Splunk. Cisco paid no less than $28 billion for a company with $3.6 billion in revenue. Juniper’s revenue is about $2 billion higher, but a price tag that’s half that of Splunk.

After the initial surprise, however, we do see several reasons why HPE acquiring Juniper Networks makes sense. We’ll list those below.

Mist AI makes Juniper very interesting

The most obvious reason why HPE is interested in Juniper Networks is the presence of Mist in Juniper’s portfolio. Mist Systems, as the company was called before its acquisition by Juniper in 2019, has had profound impact on Juniper. After all, prior to this acquisition, Juniper had no access points in its portfolio. Attempts to do so through partnerships with Ruckus, Aerohive and others did not really succeed.

Far more important than the addition of the access points per se was what was running in the cloud above those access points at Mist. Indeed, Mist’s management layer was a lot more modern than what players such as Aruba but also Cisco (Catalyst and Meraki) could offer. In particular, the platform built on microservices was already very far along in deploying AI in this environment. Thanks to AIOps, it is possible to set up, manage and generally operate (large) networks much more efficiently. With Walmart as its flagship customer (with 500,000 APs), Mist was and is definitely in the game, so to speak.

In the years since the acquisition, Juniper has added Mist into virtually every tier of its portfolio, including the popular SRX series of firewalls. So that also means integrations with Juniper’s also quite popular Junos OS. The openness of Junos in particular is appreciated we know from several conversations we’ve had about it with customers, partners and people from Juniper itself.

HPE Aruba had other concerns

Looking at Aruba, we see a slightly different picture. Over the past five years or so, they have mainly been working on the transition to SaaS, the offering that goes by the name of Aruba Central. Then there also was a lot of focus on integrating Aruba Central into HPE GreenLake. There is some AI available to Aruba customers, but it all less sophisticated than is the case with Mist. However, HPE does have a lot of focus on AIOps lately. That’s also why it acquired OpsRamp. Juniper/Mist and all the AI it brings with it could further enhance this.

Juniper/Mist approaches the access point mainly from the software perspective. HPE Aruba, on the other hand, historically had and still has more focus on the hardware. We would venture the proposition here that Aruba’s access point hardware is better than Mist’s, also based on conversations with the market we have had over the years. So merging those two environments could provide the best of both worlds. Mind you, that means yet another integration for Aruba. That won’t fill the Aruba people with much joy. It’s not ideal for Aruba’s immediate business objectives, we imagine.

Juniper has things Aruba doesn’t have

HPE has been putting a lot of emphasis in recent years on completing its networking portfolio. Last year alone, it bought an SSE player (Axis Security) and a Private 5G player (Athonet). This is to make HPE GreenLake as a platform where customers can buy basically anything (fully or partially as a service, or not as a service at all) as interesting as possible. The possible acquisition of Juniper also fits into this. This does not mean that there isn’t much overlap between Juniper’s and Aruba’s offerings, by the way. Because there definitely is. It is mainly the differences that make Juniper an interesting acquisition candidate.

HPE Aruba has traditionally focused on campus networks. In recent years, however, it has also made the transition to the data center. From Aruba Central, that’s also easy to manage. We recently spoke with Dobias van Ingen, EMEA CTO and Vice President of Systems Engineering at HPE Aruba Networking, who indicated that Aruba is the first to be able to manage a data center from the cloud. Juniper, on the other hand, is traditionally already very strong in the data center and does not focus as much on campus networks. There’s a shift in that too, of course, if only because of the Mist access points. However, there is no question that the two players together cover both environments better than they do alone.

Juniper has a strong presence in the service provider market. That also makes Juniper interesting to HPE, if it is serious about completing its portfolio. Just about a year and a half ago, Juniper announced an entirely new line for operators, Cloud Metro. This line is focused on Metropolitan Area Networks. That’s something Aruba doesn’t have.

If we keep in mind the coverage surrounding the most recent edition of HPE Discover, filling in the gaps in HPE’s (Aruba’s) networking portfolio actually makes pretty good sense. Barcelona was very much about building an AI-native architecture. That, according to HPE, is necessary for AI to run well for and within organizations. The network plays an important role in this. Ideally, HPE has all of that in-house to be able to set it up optimally. That becomes a posibility after the acquisition of Juniper.

What are the implications for companies (and for customers)?

All in all, there are good reasons why Juniper is an interesting acquisition for HPE. However, let’s also not forget that any acquisition such as this impacts the market. That is particularly the case here because it involves two obvious leaders in the networking market. In other words, a CIO who wants to acquire a new network and usually shortlists the leaders for it will have less choice after this acquisition.

On the other hand, it opens the door for new players to fill that gap. That is not easy, however, without going on an acquisition path themselves. We have seen the latter with Extreme Networks. That company and Cisco will view this possible acquisition with mixed feelings. It makes the shortlist shorter, which they will see as a positive. However, it also makes HPE much more dominant in the market. That will not be met with much cheer, even if, of course, it remains to be seen whether Cisco and Extreme will lose more deals because of this. In the short term probably not anyway, since nothing will change immediately. For the longer term, however, they will have to reckon with a more integrated proposition from HPE.

From the networking enthusiast’s perspective, we hope that Juniper does not go in the same direction as many of the other acquisitions HPE has made in the past. Not much is left of those companies other than the necessary IP (Intellectual Property). However, HPE has also shown with Aruba that things can be done differently. Let’s hope the same will be true for Juniper, should the acquisition go through.

Also read: HPE Aruba is working on “Aruba Central Next Gen”: what is that?