Watchlist: INTC, P, HPE, SNX, LNVGY
Intel (INTC) went to the market Tuesday for (another) $5 billion.
The chipmaker upsized its common-stock offering to $20 billion from the $15 billion announced a day earlier, an increase that comes as the AI buildout is forcing tech companies to secure more capital, more components and more computing capacity at once.
Intel priced the offering at $95 a share and expects about $19.7 billion in net proceeds, according to a press release issued by the company early Tuesday. Intel said the money will be used for general corporate purposes, including capex and working capital, after already having raised its 2026 capital-spending outlook to $20 billion last month.
The timing here is hard to separate from what Intel itself says it is seeing from customers.
When the initial offering was announced Monday, Intel pointed to “strong and sustainable demand” driven by unprecedented investment in AI compute. It also marked out opportunities in physical AI, purpose-built silicon, advanced packaging and external wafer manufacturing.
The company has not said that those conditions prompted Tuesday’s $5 billion increase. But, another warning from Wall Street could signal how quickly the economics of the AI buildout are changing.
Morgan Stanley analyst Erik Woodring said surging memory prices are becoming a structural problem for corporate buyers, according to a report issued Tuesday by Yahoo! Finance.
“First, it’s increasingly clear enterprises are viewing memory ‘Chipflation’ as a multi-year structural headwind,” Woodring said in a note to investors. “Rather than delaying or deferring hardware purchases until pricing cools, enterprises are quickly prioritizing/accelerating purchases of PCs, servers and storage arrays to lock in the most favorable prices and limit supply shortages, aka the ‘Fear of Missing Procurement’.”
Moving from waiting for better prices to locking down supply could keep demand elevated across servers, storage and the components that feed them. JPMorgan strategist Jay Kwon separately warned that the memory supply-demand shortage could continue for another two years.
The beneficiaries are spreading beyond the obvious chip names.
Woodring pointed to Hewlett Packard Enterprise (HPE), Everpure (P), TD Synnex (SNX) and Lenovo (LNGVY) as opportunities, although he cautioned that hardware stocks broadly have already rerated a good amount.
Wedbush also highlighted Everpure on Tuesday after the company disclosed a design win with a second top-five hyperscale customer. The firm expects a meaningful revenue contribution beginning in fiscal 2028, with gross margins between 75% and 85%. Wedbush said the win could be followed by another hyperscale customer and expects the company’s core business to exceed expectations over the next several quarters.
Elsewhere, the same pressure is showing up in the race to secure compute itself. Wedbush noted reports that Microsoft (MSFT) has discussed capacity with Taiwan Semiconductor Manufacturing (TSMC) for more than 300,000 Maia 300 chips for 2027, while Nvidia (NVDA) this week announced partnerships with six major financial institutions aimed at mobilizing more than $500 billion for AI infrastructure.
Intel’s offering fits into this larger picture because the AI boom is becoming a race for the capital and physical capacity needed to keep building.
