Noting that it was working in an period of extraordinary transformation whereby synthetic intelligence (AI) was already the bedrock for productiveness, development and effectivity, HPE CEO Antonio Neri has warned that improvement work in networking shouldn’t be progressing on the similar charge as different important know-how fields to provide the core know-how that can help the generative AI (GenAI) workloads of the long run.
As he delivered the opening keynote at HPE Uncover Europe, Neri famous that among the many firm’s core values was to have an enduring affect on society because it was fixing technological and enterprise challenges, with GenAI representing the “key shift” in the best way this was going to occur. GenAI, he stated, was not nearly speedy development – it racked up 50 million customers lower than a month after launch, whereas it took the cell phone {industry} 12 years to do the identical – however as an alternative a significant paradigm shift, driving “a profound transformation not like something that now we have ever seen”.
“AI is not only enhancing know-how, it’s enabling new worlds of interplay capabilities, relating to what is feasible in our lifetimes,” stated Neri. “It’s reworking each sector, each line of enterprise, and creating alternatives we couldn’t even think about 18 months in the past. The AI future is right here, and it’s calling us to motion. Companies that rapidly deliver AI into their enterprises won’t simply keep aggressive, they may set an ordinary.”
Neil MacDonald, govt vice-president and basic supervisor of HPE’s compute enterprise, stated: “We live on this second of super transformation the place AI is both going to raise what we do or disrupt people who don’t adapt to make the most of it. We’re dedicated to supporting our prospects whether or not they’re constructing fashions, operating companies or operating sovereign AI initiatives. And within the huge image, it’s not nearly compute: that’s [just] one of many huge modifications. It’s about bringing collectively the networking, the storage, the compute, the companies right into a seamless expertise.”
Citing a McKinsey examine predicting that as much as 30% of the hours labored immediately shall be absolutely automated by 2030, Neri noticed that GenAI had the potential to essentially rework human productiveness, which means HPE needed to be GenAI-ready, data-driven and related.
He stated hybrid cloud could be key because it allowed enterprises to course of information “the place it makes most sense”, whether or not it was on-premise, or in a non-public or public cloud.
Neri additionally introduced that HPE was introducing an choice for disconnected administration of personal cloud companies to supply a totally air-gapped administration choice. Such functionality, he harassed, was solely attainable if companies had a sturdy networking infrastructure connecting all of these factors.
Nevertheless, Neri warned of a elementary concern relating to the basic infrastructures on which these GenAI-centric, data-driven companies could be primarily based.
“As we see what is going on in [networking], whereas servers and compute [development] is shifting dramatically, the community has to catch up,” he stated. “There was an incumbency there that has seen [the sector] not innovate [recently] as quick as perhaps within the first decade and a half [of the century].
“Clients are telling us they want alternative, they want flexibility, but additionally, they want a core basis by which they will deploy these new capabilities in an accelerated method,” stated Neri. “We stated a few years in the past, the work shall be tougher. And we have been proper, and we see immediately, whether or not it’s repatriation workloads due to information sovereignty points, or due to value points, or due to AI – which by design is a data-intensive workload and hybrid as a result of information [goes] all over the place – [then the] community wants to deal with the calls for of connecting all that information to these fashions which might be being skilled, and networking is the core basis of that.”
Juniper Networks acquisition
He stated the pending acquisition of Juniper Networks, set for a while earlier than early 2025, would turn into “essential” in fixing these challenges.
“With the AI masses, information must move seamlessly between environments, and also you simply take all of them,” stated Neri. “And that’s why [Juniper] is a big plus for us as an organization, and why working with HP shall be supercharged by this acquisition, including Juniper’s high-performance, AI-native networking resolution to our security-first Aruba networking options.”
HPE acquired Aruba Networks in 2015, and since then has been constructing its networking portfolio, together with a number one campus and department providing. Neri stated that after it closes the Juniper acquisition, it’s going to in the end have an “industry-first” safe, AI-native community, objective constructed with AI and for AI, to ship a “high” end-to-end consumer and operator expertise.
Juniper Networks CEO Rami Rahim stated the acquisition was about extra than simply combining strengths, however as an alternative about reworking what networks can do for all people who use them, from delivering higher experiences with conventional purposes to assembly the rising calls for of AI workloads.
“AI is reshaping each {industry}, and to help that, networks must evolve, and so they’ve acquired to [do that] quickly,” he stated. “They’ve to alter. The objective is to simplify the lives of community operators in order that they will concentrate on higher methods to assist their prospects, designers, programmers, college students, employees, sufferers and company to make use of their networks to work, stay and play. We’ve got labored with numerous prospects to scale back issues like troubleshooting time and operational prices, and we’re getting very, very near nearly eliminating all bother tickets.
“With our know-how firms, like ServiceNow, we are able to scale back their bother ticket standing by 90% and extra,” stated Rahim. “Finally, we let IT employees concentrate on extra strategic issues. Like [motoring and cycling] retailer Halfords within the UK, who, in the course of the pandemic, noticed prospects change their buying habits, counting on on-line ordering and phone with pickup. When the pandemic eased, it capitalised on that change [and] made it simpler for in-store buyers to analysis, evaluate order merchandise from their smartphones or in-store digital shows, all powered by a single, safe community.”
He famous that Juniper has been constructing AI for networks for almost 9 years, and that with its Mist AI service it could actually automate and optimise networks in ways in which weren’t attainable earlier than, predicting points earlier than they occur, lowering working bills by as much as 85% in some situations and reducing community downtime considerably.
Seamless and scalable
As he closed out, Neri re-emphasised that prospects wanted “seamless, scalable options, with out the operational complications”, and that they’d require a community structure purpose-built for AI workloads and a number of ranges of connectivity, and with a full networking stack, together with safety.
“This isn’t nearly {hardware} efficiency and scale,” he stated. “For our prospects to in the end achieve success, they want a community stack that’s open and, very importantly, versatile – that doesn’t lock [them] in. They want us to simplify deployment and administration with automated operations that use AI ops, but additionally content-based networking. We’re prepared to assist our prospects [and] meet and sort out these AI challenges proper now.”
