Nvidia‘s (NASDAQ:NVDA | NVDA Price Prediction) legendary top boss may very well be playing a game of chess while others in the AI scene play checkers. Undoubtedly, with the Vera Rubin era on the horizon and no hesitation from the mega-cap tech giants who are expected to keep on spending mouth-watering sums on CapEx (a lot of which is going towards next-generation AI chips), it feels like Nvidia stock is nothing short of a bargain as the price-to-earnings (P/E) multiple slips below the 30 times mark for the first time in a long time.
Arguably, the Vera Rubin boom alone would be enough reason to pick up the stock as it sags below the $200 per-share level again. And while things are continuing to look up, perhaps way up, for AI demand as we enter the second half of the year, questions linger as to what could happen once custom silicon (think ASICs) looks to displace GPUs in the data centers of tomorrow.
Nvidia stock is under pressure, but it has a new growth pathway as the AI revolution matures
The hyperscalers aren’t just backing up the truck on GPUs, but they’re also spending considerable sums on the research and development of AI chips that might just help many of Nvidia’s biggest customers diversify away from the behemoth. In any case, it feels like the market is big enough that Nvidia’s shelves could be emptied and a few custom silicon players could make a move into the space.
Most notably, Alphabet‘s (NASDAQ:GOOG) Google could unlock a significant profit stream for itself as it looks to sell TPUs to firms that would have otherwise bought GPUs. In any case, the big question for Nvidia, though, isn’t just whether the firm can excel by playing defense against a number of firms that want more cost-effective chips for the inference inflection point.
With agentics and robotics on the horizon, a strong case could be made that more than just Nvidia is going to need to step up to the plate to meet that demand. And as other firms begin to make noise with their own silicon, my guess is that the cost of tokens will move lower, bringing forth even more demand. As token costs collapse and large language models (LLMs) become less large, questions linger as to what happens once AI finds its home in the edge.
Don’t discount the potential of local AI compute
Indeed, if consumers aren’t so happy to pay for AI subscriptions, perhaps making the hardware investment upfront could be the move. Add backlash and NIMBYism facing new AI data center builds into the equation, and perhaps the edge could represent the next big opportunity in the scene.
Arguably, Apple (NASDAQ:AAPL) already has a solid stage set with its latest foundation models and the architecture behind them (Instruction-Following Pruning) that allows iPhones to pack quite a punch, given the hardware constraints (a minimum of 12GB of RAM in this case).
Either way, it’s clear that Nvidia’s CEO isn’t just going to wait around as the AI revolution evolves and more people consider how they can use AI most economically. As we shift gears from tokenmaxxing towards deliberate, efficient use, I do think that the edge could be a source of tremendous positive surprises. And it’s not just about phones, either. AI PCs have had a rather sluggish take-off thus far, but that may soon change, especially as Nvidia looks to empower the PCs of tomorrow.
Nvidia’s ticket to the edge AI boom
Whether we’re talking about the RTX Spark superchip or the partnership with Microsoft (NASDAQ:MSFT), I do think that Nvidia is well-positioned to have a piece of the edge AI boom. Indeed, the Mac versus PC war could get that much fiercer with edge AI and Nvidia hardware thrown into the equation.
In any case, perhaps Jensen Huang is right on the money when he says things like AI supercomputers might be common in the home. It sounds far-fetched on the surface, but, in my opinion, the stage is already set for such with RTX Spark on the PC side and Apple and its M-series chip on the Mac side. As everyday consumers opt to use more local compute and less from the cloud, I do think that Nvidia is well-positioned to profit.
At the end of the day, Nvidia’s reach spans all major layers of what Jensen Huang refers to as an “AI cake.” And in that regard, shares seem way too cheap today, given that shares still seem priced as a cyclical GPU seller that’s nearing some sort of peak — something that I believe is far from reality.
In my humble opinion, Nvidia has what it takes to win at home and in the cloud.
Contact [email protected] for any questions or corrections.
Top Gaining Stocks
Top Losing Stocks