To The Moon

AMD’s 13.26% surge on Monday, November 9, caught investors’ attention, but the more important development may be what happens after the announcement.

Google Cloud confirmed that AMD’s EPYC processors will serve as host CPUs for its TPU 8tx systems. This gives AMD a stronger position in Google’s AI infrastructure and highlights an opportunity that investors may have previously overlooked.

For much of the AI boom, Nvidia has dominated the conversation. Investors have focused on GPUs, AI accelerators and the enormous spending by hyperscalers to build computing capacity.

But AI infrastructure is not built around GPUs alone.

As AI systems become larger and more complex, demand for server CPUs, memory, networking and storage also increases. AMD’s Google partnership is a reminder that several companies can benefit from the same wave of AI investment.

The question now is whether this strategic win can become a meaningful source of revenue.

1. Why Google’s decision is important

Google has invested heavily in developing its own Tensor Processing Units (TPUs), which are designed to accelerate AI workloads within its cloud infrastructure.

However, custom accelerators still need CPUs to coordinate workloads, manage data and support the wider computing environment.

AMD’s EPYC processors will fill this host CPU role in Google’s TPU 8tx systems.

This is strategically important because AMD does not need to replace Nvidia’s GPUs or Google’s TPUs to benefit from AI spending.

Instead, it can supply another essential layer of the infrastructure.

The partnership also provides an opportunity to strengthen AMD’s reputation with other large cloud providers. A deployment involving a major hyperscaler can demonstrate that EPYC processors are suitable for demanding, large-scale computing environments.

That could help AMD compete for additional cloud and enterprise business.

2. The bigger opportunity may be beyond traditional GPU demand

One of the biggest assumptions in the AI investment story is that the companies supplying GPUs will capture most of the financial benefits.

But hyperscalers are developing increasingly diverse computing architectures.

Google has its TPUs. Amazon has its Trainium and Inferentia chips. Microsoft is also developing its own AI silicon.

These custom accelerators are designed to improve performance and control computing costs.

However, they still operate within larger systems that require general-purpose processors.

This creates an opportunity for AMD.

If hyperscalers continue expanding custom AI infrastructure, demand for EPYC processors could grow alongside demand for their internally developed accelerators.

In other words, AMD could benefit even when customers decide not to rely entirely on Nvidia’s GPU platform.

That is an important distinction when evaluating AMD’s long-term AI opportunity.

3. The real test is whether Google orders more EPYC processors

A partnership announcement is encouraging, but investors ultimately need evidence of commercial success.

There are several developments worth watching.

Deployment expansion: Is Google using EPYC processors across a wider range of TPU systems, or is the current announcement limited to a specific deployment?

Shipment volumes: Are processor shipments increasing as Google expands its AI infrastructure? Larger and recurring orders would provide stronger evidence of demand.

Cloud revenue growth: Does AMD report continued growth in its data-centre business, supported by stronger server CPU demand?

Additional customers: Will other hyperscalers follow Google’s approach and select EPYC processors for their own AI systems?

The last point could be particularly important. One major customer is a valuable win, but several large cloud deployments would suggest AMD is benefiting from a broader industry trend.

4. Why investors should not rush to conclusions

Despite the positive announcement, there are still important unanswered questions.

AMD has not disclosed the contract value, expected shipment volumes or the timing of the financial contribution from Google’s deployment.

Without these details, investors cannot accurately estimate how much additional revenue the partnership will generate.

There is also a difference between strategic importance and immediate earnings impact.

Even if Google becomes a significant EPYC customer, the contribution may build gradually as infrastructure is installed and expanded.

Investors should also remember that AMD faces competition from Intel in server CPUs and Nvidia in AI accelerators. Hyperscalers developing their own chips may create opportunities for AMD, but they could also increase pressure on suppliers over pricing and margins.

A strong share-price reaction does not remove these risks.

5. Nvidia’s decline makes the sector rotation interesting

Another reason Monday’s trading attracted attention was the contrast between AMD and Nvidia.

AMD jumped 13.26%, while Nvidia fell 4.27%.

Some investors may see this as a sign that money is moving towards alternative AI suppliers.

However, I would not interpret one trading session as evidence that Nvidia is losing its leadership.

Nvidia still has a strong hardware and software ecosystem, while AMD is building its own position in AI computing.

The more interesting possibility is that the AI market is becoming broader.

As companies invest in different computing architectures, opportunities may spread across GPUs, custom accelerators, CPUs and networking equipment.

AMD’s Google deployment fits into this wider trend.

6. What could make AMD’s rally sustainable?

For me, the next stage of AMD’s investment story depends on three things.

First, EPYC needs to demonstrate consistent growth in the server CPU market. The Google partnership is encouraging, but broader customer adoption would make the opportunity more convincing.

Second, AMD needs to convert AI-related demand into stronger revenue and earnings. Investors should look for sustained data-centre growth rather than relying on individual announcements.

Third, profitability matters. Winning business is valuable, but the long-term benefit depends on pricing power, margins and the ability to generate cash from rising sales.

If these indicators improve together, the Google partnership could become more than a short-term market catalyst.

My take

I see Google’s decision as a meaningful strategic win for AMD because it reinforces the idea that the AI opportunity extends beyond GPUs.

The company has an opportunity to benefit from the growing demand for computing infrastructure, even as hyperscalers develop their own accelerators.

However, I would separate the excitement surrounding the announcement from the financial evidence still needed.

The next important milestones are wider deployment, recurring EPYC orders and stronger data-centre revenue.

AMD’s 13.26% jump shows that investors recognise the potential. But a sustainable rally will require more than a partnership headline.

The real opportunity is not simply getting a place inside Google’s AI infrastructure. It is turning that first deployment into a repeatable business with multiple major cloud customers.

For discussion purposes only, not financial advice.