Nvidia (NASDAQ:NVDA | NVDA Price Prediction) faces some intense competition in the AI chip scene, and that’s likely part of the reason why the shares have been flatlining since August. Of course, all it takes is one incredible, breakout quarterly earnings performance (a side of encouraging commentary from CEO Jensen Huang couldn’t hurt as well!) for Nvidia stock to start moving higher again. Undoubtedly, shares have consolidated before, and they’ve even encountered the odd painful bear market. But it’s the buyers of weakness who were rewarded, even though it felt unpleasant to get in at the time.
While Nvidia has done a fantastic job of maintaining its lead in GPUs, questions linger as to what happens once custom silicon starts to really take off. Of course, top-of-the-line GPUs will always be necessary for training frontier AI models. But as new custom silicon innovations come to be, one has to think that up-and-coming rivals might be able to chip away at Nvidia’s dominant share of the chip market.
The Mag Seven are betting big on AI inference chips. That’s more competition for Nvidia
What’s more, the Magnificent Seven companies have been making big strides with their in-house silicon. Microsoft‘s (NASDAQ:MSFT) latest Maia 200 AI inference chip is quite impressive, with reportedly big gains in performance over the third-generation Amazon (NASDAQ:AMZN) Trainium chip, as well as Google’s (whose parent company is Alphabet (NASDAQ:GOOGL)) seventh-generation TPUs (Tensor Processing Units).
Undoubtedly, the impressive results of the latest Maia chip seem to be a bigger rival to its rivals in the public cloud than the likes of the GPU titans. Still, Nvidia has been making quite a few deals in inference chips (think the Groq deal) to get ready for what could be quickly intensifying competition from ASICs.
Arguably, Jensen Huang and his team have reacted the right way as fellow Mag Seven firms have gotten more aggressive with their custom silicon ambitions, with inference at the top of mind. That said, there’s no question that companies are eager to reduce their dependence on the GPU makers as they look to take a bit of the market share away from the likes of Nvidia. Of course, it’s not going to be easy to take Nvidia’s throne. But I do think it’s a mistake to underestimate the ambitions of the Magnificent Seven innovators as they look to improve their positioning in the AI chip race.
Undoubtedly, whether we’re talking about the big three cloud hyperscalers or the likes of Apple (NASDAQ:AAPL) and Meta Platforms (NASDAQ:META), which are also hard at work on AI chips of their own in Project “Baltra” and MTIA (Meta Training & Inference Accelerator), Nvidia will need to stay on their toes to stay competitive in inference.
Could we see more pivots toward inference from Nvidia in the new year?
It’s definitely possible. The Groq partnership is definitely a big step in the right direction as Nvidia prepares to take on its Mag Seven rivals. While Nvidia is bound to retain its throne in training, at least over the foreseeable future, I’m not so sure how things will play out when it comes to inference. ASICs (Application Specific Integrated Circuits) show tremendous promise, and if the Mag Seven can lock in their cloud customers on their own silicon, there’s no telling what the implications could be for Nvidia.
At the end of the day, it’s Nvidia that’s the top performer to strive to be more like. And while Jensen Huang’s empire has everything it takes to take on more competitive forces, I do think that it’s going to be interesting to see what the next big move is as the AI chip scene mints big winners beyond Nvidia. Can the Mag Seven catch a bit of that AI chip momentum in the coming years? I’d say it’s likely, but whether that takes the wind out of Nvidia stock’s sails remains the big question.
Top Gaining Stocks
Top Losing Stocks