In short: Reports Wednesday, and Terranova sets a high bar: "this company has not grown at the pace of their revenue of 30-plus percent since the Great Financial Crisis. What does that do? That indicates you have to prove yourself." Constructive on the AI-server opportunity, sceptical the company converts it — a show-me name rather than a buy.
HPE also sells AI servers and reports Wednesday, but Terranova treats it as a show-me story rather than a buy. His reason is history: the company has not grown revenue at anything like 30% since the 2008–09 financial crisis, so an AI-driven forecast that implies it suddenly can is a claim the company has to prove, not one investors should assume.
This is the same discipline he applies to his own rotation later in the show — check whether the business has ever actually delivered the growth now being priced in before paying for it.
In short: Paired with Dell in the same market-cap roll call — "HPE is 74 billion. How did that happen?" — as evidence of how much value now sits in AI-adjacent hardware names and therefore how wide the read-through from NVIDIA's guidance runs. No stance on the company itself.
In short: Giroux: on the in-sourcing losers list (with Dell, SMCI, CAT, NVDA). Black piles on: in a hyperscaler-server comparison "Dell won hands down" — cheaper cost of ownership, easier to maintain.
In short: Q2 rev +40% to $10.7B, EPS more than doubled to $0.79; networking +148% to $2.7B on the Juniper deal; AI systems orders +$1.8B (cum. AI bookings $16.4B), $5.9B backlog. Raised FY26 guide (growth 29–33%, FCF ≥$3.5B) and pulled long-term targets in by two years. Shares +37% — biggest gain since the 2015 HP split. The gate: DRAM/NAND supply/pricing pressure could persist into 2027.
HPE sells servers, networking, and AI computing systems to enterprises and governments. This was a breakout quarter: revenue +40%, earnings more than doubled, and the stock jumped 37% — its best day since HP split in 2015. The standout was networking (+148%), supercharged by its acquisition of Juniper, plus a fast-filling order book for AI systems ($16.4 billion of cumulative AI bookings and a $5.9 billion backlog). HPE even pulled its multi-year financial targets forward by two years.
The one thing to watch is supply, not demand. HPE's machines need memory chips (DRAM and NAND), and the CEO warned that memory pricing pressure could persist into 2027 — meaning the cost or availability of components could throttle how fast HPE turns that big order book into delivered, profitable systems. So the bull case (AI demand pulling forward) is real, but the gate on it is the memory-chip supply chain.
Nothing matches this filter.
Verbatim excerpts from the public transcripts (auto-pulled at each mention's timestamp, lightly cleaned). Timestamps link into the video; "source page" opens that commentator's full analysis at this row. Click a mention's header line to expand it (one open at a time). For personal study — not investment advice.