Hewlett Packard Enterprise CEO Antonio Neri Wednesday told shareholders that competitors can’t match HPE’s “unique” AI portfolio that covers all aspects of the AI market opportunity.
“I don’t believe our competitors are ahead of us,” said Neri at the company’s annual shareholder meeting in response to a question on what HPE is doing to stay ahead of competitors who are “getting more traction” in the AI race. “I think we still have work to do to clarify a little bit on how we report our numbers because ultimately HPE has a unique portfolio that covers all the aspects of AI that we discussed before which is training, tuning and inferencing.”
Neri said the AI opportunity represents “one of the biggest opportunities” in the 85-year history of HPE which was originally part of HP, founded by William Hewlett and Dave Packard in 1939. “This is an inflection point at the right time with our technology,” he said.
Furthermore, Neri said, HPE’s $14 billion acquisition of Juniper Networks, the maker of the Mist AI networking platform, will accelerate HPE’s AI business. “By design AI is a distributed workload that needs the network and the network efficiency and the combination of what we have today in the supercomputing networking side with Juniper will give us a significant advantage for years to come.” The Juniper Networks acquisition, which must receive regulatory approvals. is expected to close late this year or early 2025.
Neri has said the Juniper acquisition, which will double HPE’s networking business, is aimed at disrupting the networking “status quo” with a new modern AI-driven networking fabric.
Here are the five “key differentiators” Neri pointed to that give HPE a competitive advantage in the AI Market:
Number 1: HPE’s ‘Supercomputing Heritage’
“Our supercomputing heritage gives us a unique position to run these massive AI systems at scale,” said Neri. “No one else has that. When it comes down to these types of leadership systems there is actually only a couple of companies left on the planet (that can produce these systems).”
In fact, Neri said, HPE next week will unveil a new breakthrough supercomputer system for one of the US government’s computing labs.
Among the prescient AI acquisitions made by Neri is the $1.4 billion purchase of supercomputing leader Cray Inc. in 2019 and the acquisition of high-performance computing powerhouse Silicon Graphics for $275 million in 2016.
HPE also acquired AI sotware startup Pachyderm in 2023 to round out its AI at scale portfolio.
HPE recently was awarded a deal for Poland’s most powerful supercomputing system for the Academic Computer Centre Cyfronet of the AGH University of Science and Technology. That system- which will be used for AI-driven research – is based on the HPE Cray EX supercomputer with the HPE Slingshot interconnect fabric with Nvidia Grace Hopper GPUs.
Number 2: Sustainability Superiority With Liquid Cooling
“As we think about this technology evolution to be able to deploy this (AI) infrastructure in a sustainable way will require new innovation particularly on the cooling side,” said Neri. “HPE has more than 500 patents when it comes down to liquid cooling which we intend to assert in the market. That gives us a position of strength that we can build on.”
Number 3 And 4: Strength In Capturing AI Service Providers And Model Builders
“What we see today with AI action is pretty much in two buckets: service providers and what I call the model builders,” said Neri. “We are capturing a significant opportunity in the market as we reported last quarter in our order bookings and we have a significant pipeline ahead of us.”
In the most recent quarterly earnings call, Neri told analysts that AI server demand remains strong characterized by “growing” order bookings.
HPE has combined the compute and HPC and AI segments into a single server segment that integrates supercomputing, high performance computing, general purpose computing and AI systems. “This will enable us to maximize the opportunities across the entire AI lifecycle from training to tuning to inferencing and execute with agility,” said Neri in the last quarterly earnings call.
Neri said at the time that among the wins in AI inferencing is a deal with Coles Supermarkets, an Australian retailer, that recently implemented an HPE ProLiant Gen11 AI inferencing solution, which helps with video imaging to reduce store stock lost to theft and incorrect scanning at the checkout.
Number 5: Services Capabilities To Tune And Run AI Systems
“Last but not least, which is super important, when it comes down to running these systems you need a lot of services capabilities,” said Neri. “When you look at some of our traditional competitors they don’t have that. That is something that will help us continue to grow because most of these systems customers are not going to run themselves because they don’t have the data center space, they don’t have the cooling, they don’t have the power, which is one of the biggest challenges today and we actually know how to run the systems. In fact, we have been building a couple of sites where customers are actually deploying the infrastructure with us and we are running it for them.”
Michael Maher, director of professional services at CPP Associates, Clinton, New Jersey, the GreenLake North America Partner of the Year, said HPE is far ahead of competitors in the AI race.
“HPE has the pedigree, the history and the forward-looking business plan to carry AI forward,” he said. “Everybody else in the market is playing catch up to HPE. Listen, HPE holds world records for supercomputing, they have robust deep ties with key partners like Nvidia, and HPE has all of the vital infrastructure components and expertise to bring customer’s AI visions and requirements to fruition. We are having daily discussions with customers about how we can use HPE technologies and our shared expertise to get our customers AI initiatives launched, operational and yielding results.”
