To The Moon

$NVIDIA(NVDA)$  leadership uncertainty is creating an interesting shift in the semiconductor market. While Nvidia shares plunged around 9% in after-hours trading on Wednesday, $Advanced Micro Devices(AMD)$  gained more than 4% and $Broadcom(AVGO)$  advanced over 2%.

Investors appear to be looking beyond Nvidia and considering which companies could benefit if customers begin diversifying their AI chip suppliers.

But AMD and Broadcom are not competing with Nvidia in exactly the same way. Each has a different position in the AI infrastructure market, and both could benefit from the next phase of AI spending.

1. AMD: The direct challenger to Nvidia

AMD is one of the most obvious alternatives for investors looking for exposure to AI accelerators.

Its Instinct GPU series targets data-centre workloads, including AI training and inference. As cloud providers and large enterprises expand their computing capacity, AMD has an opportunity to win customers that want alternatives to Nvidia.

One important factor is supplier diversification. Technology companies may not want to depend entirely on one chipmaker, particularly when AI infrastructure has become strategically important.

However, AMD still faces a significant challenge. Nvidia has an established advantage through its CUDA software ecosystem, developer support and relationships with major cloud providers.

For AMD, the opportunity is not simply to sell competitive hardware. It must also convince customers that its products can deliver reliable performance, software compatibility and sufficient supply at scale.

2. Broadcom: A different way to benefit from AI spending

Broadcom has a different business model.

Rather than competing primarily through general-purpose GPUs, it develops custom AI accelerators, known as ASICs, for major technology companies.

These chips are designed for specific workloads and can offer advantages in performance and power efficiency in suitable applications.

Broadcom also supplies networking technology that connects processors and servers across large data centres.

This is important because building AI infrastructure requires more than powerful processors. Companies also need high-speed networking to move enormous amounts of data between computing systems.

As major cloud providers develop their own AI chips, Broadcom could benefit from increasing demand for custom silicon and networking equipment.

3. Nvidia’s leadership uncertainty creates an opening, but not an immediate replacement

Jensen Huang has led Nvidia since co-founding the company in 1993. His leadership has been central to its transformation into a dominant AI chip supplier.

Recent media reports about allegations involving Huang and the board’s consideration of his future have introduced uncertainty. These remain reported allegations and speculation, not established findings or confirmation that he will leave the company.

For investors, the bigger question is whether any leadership transition could affect Nvidia’s product roadmap, customer relationships and competitive strategy.

However, Nvidia’s technology, engineering teams and software ecosystem remain significant strengths. Even if customers reconsider their supplier relationships, switching AI infrastructure is expensive and takes time.

AMD and Broadcom may gain opportunities, but replacing Nvidia’s market position will not happen overnight.

4. What should investors watch next?

I would focus on three areas:

• AMD: New data-centre GPU customers, Instinct shipments and growth in AI-related revenue.

• Broadcom: Custom AI chip orders, customer commitments and networking revenue.

• Nvidia: An official board announcement, leadership clarity and whether its business outlook remains unchanged.

I would also watch whether the recent share-price movements develop into a sustained trend. One strong after-hours session does not prove that customers are moving away from Nvidia.

My investment view

I see AMD and Broadcom as two different ways to participate in the AI infrastructure opportunity.

AMD offers more direct competition in data-centre GPUs, while Broadcom benefits from the growing demand for custom AI chips and networking.

Nvidia’s current situation may accelerate discussions about supplier diversification, but the long-term winners will be determined by product performance, customer adoption, revenue growth and profitability.

Rather than assuming Nvidia’s problems automatically mean AMD and Broadcom will win, I would look for evidence of actual customer orders and stronger earnings.

The AI market may be large enough to support several winners. The next stage could be less about one company dominating everything and more about how different chipmakers capture different parts of the growing AI infrastructure market.

For discussion purposes only.