**Space: The Next Frontier for AI Compute Power**
Tony Wang, the longtime manager of T. Rowe Price’s $16 billion Science & Technology Fund, is looking skyward. Lately, he’s zeroing in on space as the next big frontier for artificial intelligence infrastructure.
He’s convinced that the surging demands of advanced AI systems could soon push Earth’s resources to the brink. It’s a bold shift in strategy, but maybe not that surprising if you’ve watched the breakneck pace of AI.
The Growing Compute Conundrum
Wang’s main point? We’re on the verge of needing “about a hundred times more compute” than what’s out there today. That’s not just a little more server space—it’s a mind-boggling leap.
Addressing Earth’s Limitations
Earth only has so much land, and giant data centers take up a lot of it. Wang figures we’ll have to start putting some of that compute power into orbit, literally expanding beyond the planet.
Strategic Investments in the Space-AI Nexus
He’s already making moves. The fund holds positions in companies that could ride this wave, like SpaceX and AST SpaceMobile.
These aren’t just moonshot bets—they’re early stakes in what might become a whole new industry if AI’s appetite for compute keeps growing.
The Potential Impact of SpaceX’s IPO
SpaceX’s rumored IPO is a big deal. If it happens, and the company lands in major stock indexes, portfolio managers everywhere might have to shake up their holdings.
That could mean some shifting away from the usual tech giants—Microsoft, Meta, Alphabet, Apple, Nvidia—as investors look to get a piece of SpaceX’s action.
Decoding the Bottlenecks: Memory and Optical Networking
Wang’s made a name for himself by spotting tech trends early—Nvidia, AMD, Lam Research, Micron, you name it. Now, he’s turning his attention to memory and optical networking, the real choke points for AI right now.
High-Bandwidth Memory (HBM DRAM): The Engine for AI Workloads
He’s especially bullish on High-Bandwidth Memory (HBM DRAM). Companies like Micron, Samsung, and SK Hynix are the big suppliers here.
This kind of memory is essential for chewing through the massive calculations that modern AI demands.
The Ripple Effect of HBM Shortages
There’s a real shortage of HBM DRAM at the moment. That crunch is spilling over into related areas like flash storage and cold-storage—think Sandisk, Kioxia, Western Digital, Seagate.
It’s all tangled together, and a hiccup in one spot can throw off the whole supply chain.
Silicon Photonics and Optical Networking: Accelerating Data Transfer
Wang’s also pretty optimistic about silicon photonics and optical networking. These technologies help move data insanely fast and keep data centers from overheating as they get bigger and more power-hungry.
Key Companies in Optical Networking
His fund owns shares in Ciena, Lumentum, Coherent, and has a private stake in Lightmatter. They’re all working on making data transmission quicker and more efficient.
The Rise of Agentic AI and Its Compute Demands
Another thing on Wang’s radar: “agentic AI.” Instead of waiting for a human to click or type, these AI agents go out and do stuff on their own.
That’s a huge shift, and it’s sending compute needs through the roof. The cost of using and developing these models—what they call “token spend”—is ballooning, too.
Distinguishing Signal from Noise: A Value Investor’s Approach
Some companies have already slowed down or dropped AI projects because costs got too high or early tests flopped. Still, Wang sees all this ongoing experimentation and spending as a sign that valuable use cases are just starting to emerge.
He tries to set his tech-focused fund apart from big indexes like the Nasdaq-100. He wants to find companies where AI is actually driving real, lasting growth—not just hype that fizzles out.
Broader Market Optimism and AI Infrastructure
There’s a growing sense of optimism across the broader market, especially around companies set to benefit from AI infrastructure. Analysts expect strong earnings-per-share growth for these S&P 500 firms, possibly as soon as 2026.
This trend really backs up Wang’s strategic perspective.
Here is the source article for this story: Why this fund manager who was an early proponent of Nvidia is now betting on the AI space frontier