Competition Is Closing In on Nvidia

Nvidia is cutting its chip development time in half…

Until now, Nvidia released a new GPU (Graphics Processing Unit) every two years.

Companies use GPUs to train ChatGPT and other AI systems as well as power data centers and high-performance computers.

Releasing a new GPU every year may seem like a good move.

But it’s actually an admission of something I’ve been warning you about…

Nvidia’s competition is closing in.

This spells trouble for Nvidia shareholders.

After the 211% jump in its stock price this year. Nvidia is priced for perfection. And losing market share to competitors is far from a perfect outcome for the chipmaking giant.

But it’s also great news for the company I call the “next Nvidia.”

It’s about to release its own AI-ready GPU. And it will be more than twice as powerful as Nvidia’s top AI chip.

Even better, this Nvidia rival is a bargain compared to Nvidia.

But before we get to Nvidia’s downfall, you must understand how it became top dog in the first place.

Pushing the Boundaries

Nvidia’s breakthrough came in 2017. That’s when it launched its V100 GPU.

The company designed it with AI in mind.

The V100 introduced NVLink technology. This allowed multiple GPUs to interconnect into a unified computing platform which meant developers could scale up their computing power.

And it slashed the time it takes to train ChatGPT, Bard, Claude, and other AI chatbots from weeks to days.

It also allowed for better image recognition for autonomous driving. These breakthroughs would not be possible without the AI-ready computing power NVLink provided.

Nvidia again pushed boundaries in 2020 with the release of the A100. This new version of the V100 delivered further leaps in speed, scale, and flexibility.

Between the V100 and A100, Nvidia cemented an unrivaled dominance in AI chips. And the company retained its place as the premier AI chipmaker in 2022, with the launch of its new H100 GPU.

But this two-year upgrade cycle opened opportunities for competitors.

For example, Intel is developing its Falcon Shores GPU, with a scheduled release in 2025. It’s aimed at high-performance computing and AI workloads.

And it’s targeting the same customers Nvidia sells its H100 to – supercomputing centers, enterprise data centers, cloud service providers, and AI companies.

Other competitors are slated to release GPUs sooner than that.

That’s why Nvidia is shifting to releasing new AI chips annually instead of every two years. It knows the chip race has entered a new phase.

And it wants to stop its lead narrowing as rivals rapidly advance.

Wall Street analysts will try to spin Nvidia’s transition to an annual chip upgrade cycle as a sign the company continues to innovate.

But this rosy portrayal misses a more troubling scenario for Nvidia.

The Next Nvidia

Nvidia will need to make key design trade-offs to release chips at a faster pace.

For example, it’s squeezing its new B100 chip into a 2024 release.

Nvidia’s H100 chip clocks in at 700 watts (W) of power. Originally, the company intended the B100 to support a power-hungry performance of up to 1,000W. But for the initial B100 rollout next year, Nvidia is limiting power to 700W.

That’s a problem because higher power is needed to train large AI models, which can have billions of parameters. It’s also needed to run AI systems, such as in self-driving cars, in real-time.

So, developers are unlikely to want to upgrade from the H100 to the B100 because it likely won’t offer enough of a performance improvement.

And unfortunately for Nvidia, the way forward in tech is to innovate in ways competitors can’t.

That’s bad news for Nvidia shareholders. The stock’s rich valuation requires accelerating revenues and massive profits.

And sacrificing performance by rushing out new chips isn’t the way to ensure that happens.

But this is great news for folks who own shares in a rival chipmaker set to become the leader in the industry.

It trades at a price-to-sales ratio of 7.9 versus Nvidia’s price-to-sales ratio of 35. Its new AI chip outperforms Nvidia’s H100. And it’s the main reason Nvidia is scrambling to release chips at a faster rate.

It’s all in the new presentation I’ve put together for you and your fellow readers. As you’ll see, this smaller company will power the next wave of AI profits.

And it’s caught the eye of legendary trader Ken Griffin… hedge fund billionaire David Tepper… and value investor David Einhorn.

It’s only a matter of time before everyone starts piling in… because this company is about to unveil the most powerful AI chip in history.

To get ahead of the pack go here to find out more.

Regards,

Colin Tedards
Editor, The Bleeding Edge