Nvidia (NASDAQ:NVDA | NVDA Price Prediction) may very well be the king of the so-called “AI Factories,” but the GPU titan also stands to gain as the AI revolution makes its way to the edge as well. Of course, there seems to be no slowing the great AI data center buildout.
The buildout could continue to accelerate, gobbling up all of the DRAM, NAND, GPUs, and server racks, even as supply looks to make a bid to catch up to the unprecedented demand. Add water and electricity usage into the equation, and it’s clear that some structural forces are in play that could continue driving up the price of electrical components.
The solution may very well lie in the rapid advancement of on-device AI. Indeed, not every personal prompt to manage one’s email needs to go to the cloud to run on one of the most advanced AI models out there (let’s say something Mythos grade).
Nvidia’s got its ticket to the edge AI race
As routing improves and more consumers look to buy smartphones and computers specifically for running AI applications, questions linger as to just how large the opportunity in the edge could be. You’ve probably heard about the smartphone supercycle or the rise of “AI PC” for some number of years now.
Thus far, it hasn’t really materialized in a way that investors have expected. And while iPhone sales over at Apple (NASDAQ:AAPL) have been strong, it’s hardly anything that one could describe as a “supercycle” or anything close to it.
With recent price increases due to the AI-induced RAM-magadden, questions linger as to whether consumers will be willing to pay up for that so-called “AI tax” for the next generation of devices. Either way, it’s just highly unusual for the firm to increase prices on iPads or Macs in the midst of a cycle. I guess that’s just how unprecedented the situation is.
In any case, Nvidia has set its sights on the AI PC. And while the rise of edge AI might be a bit late to the party, I do think that it’s about to finally show up. As firms get “smart hybrid routing,” I do think that we could be moving on from the era of “tokenmaxxing” towards one of tokenminning,” as the right model is picked for the prompt. Any way you look at it, all of this bodes very well for the AI PC, which may finally be ready for prime time after experiencing limited success in the past two years.
Is the AI PC ready to get going?
Microsoft (NASDAQ:MSFT), the innovator behind Copilot and AI PCs, is about to get a catalyst with Nvidia’s RTX Spark superchip.
It’s not just another powerful chip. Rather, it’s something that could empower on-device agentic AI and change the game entirely. In a prior piece, I noted that with RTX Spark, Nvidia has pretty much punched its ticket to the edge AI race and that it could make Jensen Huang’s empire the king of AI, regardless of where models are run.
As AI data centers scramble to get around bottlenecks (not enough RAM or NAND to go around, it seems), I do think that making the most of limited resources — and that includes incorporating AI PCs and smartphones — could be key to advancing the AI revolution without having to sit around, waiting for those massive data centers to come online as they power through various hurdles standing in the way of getting AI compute to where it needs to be.
The TAM on the edge could be considerable
For Nvidia, the big question is how large the total addressable market (TAM) could be as it finds its way inside the next generation of AI PCs. The most advanced RTX Spark chips could be a hot seller among professionals seeking to invest in an AI workstation to keep that token bill to a minimum. Add everyday consumers into the equation, and I do think that an AI PC supercycle could be the next theme that helps Nvidia get going again.
Whether it’s supercomputing in the home or AI superclusters in the data center, it looks like Nvidia is positioned to win once again. And, in my view, that makes Nvidia stock look like a solid deal at $200 per share.
Come the fall season, I think we’ll learn more about the true long-term potential of RTX Spark and products like it as they hit the consumer market. Given consumer hardware prices only seem to move higher, I certainly wouldn’t rule out the potential for “panic-selling,” especially if it becomes more evident that RAM shortages and all the sort are more structural than cyclical.
Contact [email protected] for any questions or corrections.
Top Gaining Stocks
Top Losing Stocks