Processors burn time and power waiting on data. In a new free presentation, Tech Expert James Altucher explains the fix no one is talking about.
Washington, D.C., Nov. 03, 2026 (GLOBE NEWSWIRE) — The most powerful AI chips on Earth have a strange problem. They’re hungry.
They can run calculations at staggering speed. But they keep stalling while they wait for data to show up. Now even the companies building AI’s fastest memory are working on a different kind of fix.
SK Hynix, a leading maker of the high-bandwidth memory inside AI systems, has been pushing technologies that move storage and data closer to the processor. The message is clear. Faster chips aren’t enough. The data has to get there faster too.
In a new free presentation, tech analyst James Altucher argues this same problem sits at the center of Elon Musk’s plan for AI. “Elon built the brain,” Altucher says. “But he did not build the nervous system.”
To be clear, SK Hynix isn’t the company Altucher is pointing to. The idea is the same: AI is only as good as its ability to move information.
A Fast Brain Still Has to Wait
Picture a brilliant worker who has to walk across the building every time they need a file. The worker is fast. The walk isn’t.
AI hardware has the same problem. Processors run billions of calculations. But shuttling data back and forth between computing and memory burns time and power.
Better memory helps. But the industry keeps hunting for smarter ways to place and connect the parts. That’s why a top memory maker is chasing more than another speed bump. It’s rethinking how the whole system fits together.
Now Put the Computers Hundreds of Miles Apart
Inside a data center, engineers control everything. The racks. The cables. The networking gear.
Now imagine computers spread across space, orbiting the planet. That’s the future Altucher believes Musk is building. He points to a SpaceX filing with the FCC that lays out plans for one million AI data center satellites.
Altucher frames the stakes in sweeping terms. “Imagine if a SINGLE person controlled the entire internet from its inception,” he says. That’s the kind of control he believes Musk is going after in AI.
But launching the hardware is only step one. Those machines would have to trade data fast, nonstop and on a tight power budget. In space, the communications problem could be just as hard as the computing problem.
The Least Glamorous Corner of AI
That’s where Altucher points his audience.
Everyone watches the famous chipmakers. Far fewer watch the companies that move the data. Altucher says a brain on its own can’t do much. “A brain can’t communicate what [it’s] thinking on its own,” he says.
His free presentation focuses on one outside supplier he believes helps provide that nervous system for Musk’s satellites. He calls it a publicly traded company “that almost nobody has heard of.” And he says it’s now producing chips “at a rate of 5 million per day.”
The lesson from SK Hynix is simple. AI can’t get smarter just by thinking faster. It has to remember faster. Retrieve faster. And communicate faster.
About This Presentation
Behind every AI breakthrough is a pile of hardware nobody talks about. In this free presentation, James Altucher digs into that hidden layer, from the chips that think to the systems that carry their data. He explains why he believes Elon Musk’s orbital AI plan depends on a supplier most investors have never noticed. Watch it here.
About James Altucher
James Altucher has been managing money and building trading systems on Wall Street for three decades. He has run hedge funds, raised more than $100 million for a venture capital firm and built AI trading algorithms to compete with the very best. He’s also known for an early call on the very chips that power today’s AI boom. He first wrote about NVIDIA in 2008, calling it a “Super Chip Maker.” Viewers may know him from CNBC, Fox Business or The Wall Street Journal.
CONTACT: Derek Warren Public Relations Manager Paradigm Press Group Email: dwarren@paradigmpressgroup.com
