When HPE reported its second-quarter results, the increased adoption of artificial intelligence (AI)-based systems was one of the company's big wins.
The company reported second-quarter revenue increased 3% to $7.2 billion.
The new AI servers are expected to be delivered in just six weeks after the company resolves its supply chain issues.
In a blog discussing the results, CEO Antonio Neri spoke about the growth of AI, particularly among HPE's enterprise customers: “Demand for HPE's AI products is accelerating, and our solid execution led to AI systems revenue more than doubling quarter-over-quarter to more than $900 million, driven by supply chain transformation driven by GPU improvements.” [graphics processing unit] “Availability,” he said.
Neri added that HPE's AI business doubled from the previous quarter due to strong bookings and improved conversion rates from the supply chain.
“We have tripled the number of enterprise AI customers year over year and they now represent more than 15% of our total AI bookings of $4.6 billion,” he said.
In a blog post, Neri spoke about HPE's leadership in water-cooling technology for AI servers and its strengths in green IT. “HPE currently owns four of the top 10 fastest supercomputers in the world, all of which are direct water-cooled,” he said. “We're also proud to say we've built seven of the top 10 most energy-efficient systems in the world.”
Liquid Cooling
Regarding liquid cooling technology, he said that the liquid cooling capabilities of the company's AI servers have resulted in over 300 HPE patents.
The company also experienced strong demand for its GreenLake products, helping it exceed $1.5 billion in annual recurring revenue (ARR).
“While AI had the highest ARR growth during the quarter, the two fastest growing components of ARR, storage and networking, both maintained strong growth rates,” Neri said.
During the earnings call, Neri was asked about enterprise adoption of HPE's AI servers. According to a transcript posted to Seeking Alpha, enterprise customers are showing increasing interest in HPE's AI capabilities. “As these efforts continue to progress from the research and discovery phase, we expect to see further acceleration in orders for enterprise AI systems by the end of the fiscal year,” Neri said.
Supply chain constraints due to GPU demand are curbing many IT leaders' appetite to build out their enterprise AI strategies.
Neri also spoke about improvements to the lead time for supplying AI servers: “We're very pleased with our ability to deliver systems quickly. [within] “Six to 12 weeks,” he said.
Chief Financial Officer Marie Myers added: “In servers, we expect sequential revenue growth through fiscal 2024 driven by improving supplies of GPUs for AI systems and increased demand for traditional servers.”
Along with growth opportunities in AI, Neri said in his blog, HPE is also seeing signs of market recovery in traditional and cloud infrastructure markets. “Traditional server orders increased quarter-over-quarter and year-over-year from enterprise, public sector and small business customers in North America and Europe. Revenue also increased quarter-over-quarter as customers transition to HPE ProLiant Gen 11 servers,” he said.
The company's GreenLake offering is also benefiting from the growing popularity of AI among HPE's enterprise IT customers: The company recently announced HPE GreenLake for file storage capabilities, which Neri said includes options tailored to AI's unstructured data needs.
