Intel Corp (INTC) caught a bid on Tuesday after announcing a collaboration with SoftBank Group Corp's (SFTBY) subsidiary, Saimemory. The partnership is all about building the memory technology that tomorrow's AI systems will need—assuming those systems don't become sentient and decide they'd prefer a different kind of memory altogether.

The alliance, dubbed the "Z-Angle Memory program" (ZAM), focuses on developing memory solutions to keep pace with the exponential demands of artificial intelligence and high-performance computing. Think of it as preparing for a future where computers need to remember exponentially more things, exponentially faster.

The timeline is refreshingly realistic: prototypes should arrive by 2027, with full commercialization targeted for 2030. That's actually not too far off in semiconductor time, where "next generation" can sometimes mean "maybe next decade."

Building Better Memory for Hungry AI Systems

Intel's bringing serious credentials to this partnership, particularly its work with the U.S. Department of Energy's Advanced Memory Technology program. That initiative has focused on boosting the performance and power efficiency of next-generation DRAM for computers and servers—exactly the kind of foundation you'd want when tackling AI's insatiable appetite for memory bandwidth.

And that appetite is real. AI-related memory demand is surging, creating supply shortages across the industry. The ZAM program aims to address those bottlenecks with energy-efficient solutions that the computing world desperately needs. Intel's new memory architecture and assembly approach promise better performance while cutting power consumption and costs, which could make advanced memory more accessible once the 2030s roll around.

Meanwhile, Back at Earnings Season

The memory partnership news comes as Intel works through a bumpy turnaround. The company reported mixed fourth-quarter results on January 22, beating profit expectations but delivering guidance that left investors wanting more.

Intel posted $13.67 billion in revenue, topping the $13.37 billion consensus. Adjusted earnings came in at 15 cents per share, crushing expectations of 8 cents. But here's the catch: revenue still fell 4% year-over-year. Growth in Data Center and AI was real, but weakness in the Client Computing Group offset those gains.

The forward-looking picture got cloudier. For the first quarter, Intel forecast revenue between $11.7 billion and $12.7 billion—below the $12.49 billion analysts expected. Instead of the modest profit Wall Street anticipated, Intel predicted breakeven adjusted earnings. The stock took a beating on that news, highlighting how supply constraints and margin pressures continue to weigh on the company's outlook.

What Analysts Are Saying

Needham analyst N. Quinn Bolton zeroed in on chip supply issues, particularly with Intel 10 and 7 process nodes, which continue limiting shipment volumes. Bolton also noted that Intel's adjusted gross margin was headed for 34.5% at the midpoint, pressured by a less favorable product mix.

Benchmark analyst Cody Acree saw the stock pullback as an overreaction to cautious guidance after what was actually a solid quarter. He emphasized that Intel is navigating a major transition, with critical milestones like the 18A process and the early Core Ultra Series 3 launch on the horizon.

JP Morgan analyst Harlan Sur attributed the weak guidance to ongoing wafer capacity constraints, especially on Intel 10 and 7, combined with that unfavorable product mix everyone keeps mentioning. Sur expects Intel to streamline its server roadmap to better compete with Advanced Micro Devices, Inc (AMD), and remains bullish on Data Center and AI demand as the long-term growth driver—even as PC shipments decline.

INTC Price Action: Intel shares were up 0.45% at $49.03 at the time of publication on Tuesday, according to market data.