Alphabet Google JHVEPhoto / iStock Editorial via Getty Images

  • Alphabet (GOOG) launched Gemini 3.0 to strong reception and prompted OpenAI to enter code red mode.

  • Meta Platforms may purchase a significant amount of Google’s TPUs by 2027.

  • Google’s TPU chips could disrupt Nvidia’s GPU dominance through lower costs and power efficiency.

  • If you’re thinking about retiring or know someone who is, there are three quick questions causing many Americans to realize they can retire earlier than expected. take 5 minutes to learn more here

It’s the trillion-dollar question that many investors have surely been pondering in the past few weeks. With Alphabet (NASDAQ:GOOG) making massive strides while AI chip titan Nvidia (NASDAQ:NVDA) is stuck in a bit of a rut, questions linger as to whether we’re about to get a new AI leader and perhaps a new largest company on the planet.

Undoubtedly, Google’s Gemini 3.0 launch went well, really well. And with smart tech leaders such as Marc Benioff praising the new model, I don’t think investors should discount the leap that Gemini 3.0 made over the competition. In any case, the new model was so good that it even has OpenAI and Sam Altman entering some form of “code red” in an attempt to catch up before more users, like the great Marc Benioff, look to make the jump from ChatGPT to Gemini.

Though Marc Benioff has said he won’t be going back to ChatGPT, I do think that if Altman and OpenAI can respond effectively in the coming weeks and months, there is an opportunity to win back the many that may have made the switch since Gemini 3.0 landed. For now, it seems like the odds are tilting towards Google’s favor as it looks to run away with the lead in large language models. The benchmarks really do speak for themselves!

Beyond the momentum behind Gemini 3.0, though, is the bright light on Google’s TPUs (Tensor Processing Units) after it was announced that Meta Platforms (NASDAQ:META) might be looking to buy a considerable amount of TPUs from the firm by 2027. For now, it’s not clear why Meta Platforms is exploring such a potential deal. It is significant and could mean massive things for Google, perhaps at the expense of Nvidia, which might face a trickier time as new rivals take on AI chips from a different angle.

Undoubtedly, Nvidia’s GPUs aren’t going to be replaced with TPUs overnight. However, there is disruptive potential there, and, as I noted in my prior piece, I would much rather be in an AI chip disruptor than a firm that could face greater competition and a potential erosion of its dominance in the coming years.

Story Continues

Though I expect GPUs will continue to pave the way for further gains in Nvidia shares, I would have to side with Google and Alphabet if I had to pick a long-term winner to come out on top in the AI chip war over the next decade. While Nvidia isn’t going anywhere and is likely to stay on top in GPUs, I do think that TPUs are coming into their own.

And if Google ends up winning the AI model race with Gemini, I do think TPUs might gain even more momentum, given there’s potential efficiencies to be had by having the same firm make the models as well as the chips they run on. Also, given how expensive GPUs are and investors’ newfound distaste for heavy AI spending, perhaps AI bubble fears will push for more cost-effective solutions.

Whether that’s TPUs or in-house ASIC chips that are put in the place of GPUs, I think there are great uncertainties when it comes to who ends up winning the war of AI chips. Either way, I think the race is getting closer. And while the torch may not have been passed yet since Nvidia still claims to be “generations” ahead, I would really worry about TPU deals and the AI momentum that’s running behind Google these days.

While I could be wrong, I think the trend in AI chips will keep moving more towards cheaper, power-efficient, more customized hardware, rather than a powerful “Swiss army knife” kind of solution.

Perhaps Gemini 3.0 on TPUs won’t just be effective and fast, but it might be magnitudes cheaper than alternative products that run on GPUs. Time will tell. Either way, I think the Gemini 3.0 and Google’s AI model leadership bodes really well for how it’ll fare in AI chips as it continues to advance its incredible TPUs.

In short, while I favor Alphabet over Nvidia, there’s no reason why both stocks can’t be big winners in the new year.

You may think retirement is about picking the best stocks or ETFs, but you’d be wrong. Even great investments can be a liability in retirement. It’s a simple difference between accumulating vs distributing, and it makes all the difference.

The good news? After answering three quick questions many Americans are reworking their portfolios and finding they can retire earlier than expected. If you’re thinking about retiring or know someone who is, take 5 minutes to learn more here.