AMD's hidden AI weapon may finally be exposed

AMD's artificial intelligence story has mostly been judged against Nvidia's dominance in graphics processing units.

That's maybe too narrow.

BofA Global Research lifted its price target for Advanced Micro Devices (AMD) to $620 from $550, saying the next phase of AI infrastructure may not rely solely on GPUs. As agentic AI workloads generate new bottlenecks within data centers, central processing units are becoming increasingly crucial, the firm argues.

The target represents almost 11% upside from AMD's price of $557.89 at the time of the report. That is hardly a moonshot call.

It's a more important kind of call: a bet investors may be underestimating how much AI demand may increase AMD's server CPU business while its GPU roadmap ramps.

That includes AMD's next-generation Venice CPU, MI455X GPU, and Helios rack-scale solution, which BofA said it expects to benefit from strong EPYC server CPU demand and increased supply visibility.

That's why AMD's Advancing AI event on July 23 is so important.

Investors are now asking more than whether AMD can catch Nvidia in GPUs. The question is whether AMD can utilize its CPUs, GPUs, and rack-scale systems to be a more complete AI infrastructure supplier.

BofA said AMD is well positioned because of its CPU and GPU demand, supply visibility, and upcoming AI product launches.

AMD's AI story is expanding beyond GPUs

One issue dominates the AI chip market: who can beat Nvidia?

For AMD, the question has traditionally been about GPUs. Nvidia (NVDA) is still the king of AI accelerators, and AMD has been striving to establish that its Instinct portfolio can grab major hyperscale demand.

But BofA's report highlights another aspect of AI's system.

In agentic AI, CPUs are responsible for such things as tool calls, code execution, orchestration, memory management, and system coordination. If those operations lag, pricey GPUs may just lie idle.

That gives AMD a far larger potential.

The company already has a strong franchise in server CPUs with EPYC. If artificial intelligence systems need more and more CPU content with each generation, then AMD's server business could be a bigger portion of the AI trade than investors who just think about GPUs understand.

BofA currently anticipates the server CPU market growing to almost $170 billion by 2030 from around $35 billion in 2025. Over time, the company anticipates AI-related CPUs to make up most of that market.

Related: 5-star analyst sends AMD stock investors a warning

That's the subliminal message for investors.

And the AI story doesn't have to be just about AMD gaining share in GPUs. It might also help if the AI data center design itself becomes more CPU-heavy.

That is why the launch in Venice is so valuable.

BofA anticipates AMD's sixth-gen EPYC Venice CPUs to prolong the company's high-core-count advantage, which may be a big deal for agentic artificial intelligence systems requiring more parallel computing and higher rack-level throughput.

Agentic AI may turn CPUs into a new bottleneck

The CPU discussion is more critical than ever as AI workloads evolve.

BofA frames the battle as a debate between high-frequency CPUs and high-core-count CPUs.

High-frequency CPUs can help reduce latency. This is important when artificial intelligence systems need to make tool calls quickly. High core count CPUs can do more work concurrently, which is significant when systems are performing many agentic tasks at the same time.

BofA believes both theories are pointing to the same conclusion: artificial intelligence systems would likely require greater CPU power.

More AI:

• The new Chinese AI model rattling U.S. tech investors
• Anthropic restores access to Mythos 5 for select organizations
• SoftBank CEO offers stinging critique of Musk's AI bet

That's promising news for AMD.

AMD's current Turin CPUs support up to 192 cores, and the future Venice series will support 256 cores, the company adds. That compares to Nvidia's future Vera CPU, which will have 88 cores, and Intel's existing Granite Rapids series, which has 128 cores, according to BofA.

The number matters because power, space, and system efficiency increasingly constrain data centers.

More CPU cores inside a fixed power envelope can improve throughput. In AI systems, that can help reduce wasted GPU time and increase overall rack performance.

That's where AMD's AI story gets more fascinating.

The company doesn't just sell parts. It is attempting to sell the redesign of the data center system.

AMD's July AI event may reset the stock debate

BofA sees a good Q2 setup for AMD, but the most important element of the call may be what happens next.

The company's formal second quarter expectations are close to consensus. BofA expects revenue of roughly $11.28 billion and non-GAAP earnings of $1.62 a share for the June quarter, mostly in line with the consensus.

So the story here is not just that BofA is seeing a large second quarter beat.

BofA believes AMD's guidance, product commentary, and 2H AI plan will be more important than the reported quarter itself.

This is a crucial difference.

BofA anticipates that the third-quarter projection will begin to reflect the first MI455X Helios rack shipments ahead of a bigger fourth-quarter ramp. The company is also expecting AMD to use its Advancing AI event to feature Venice CPUs, Verano CPUs, the MI455X GPU, and the Helios rack.

The event could be AMD's chance to rewrite the story.

Instead of pitching itself only as an AI GPU challenger, AMD can show investors a broader platform: high-core-count CPUs, AI accelerators, and rack-scale systems aimed at hyperscale customers.

That's why the $620 target counts.

It's not just the near-term results driving the price-objective hike. AMD deserves a higher multiple, as the growth opportunity in CPUs and GPUs could be larger than previously assumed, BofA said.

BofA now bases its AMD price objective on 47 times expected 2027 non-GAAP earnings, up from 42 times previously.

AMD's cleaner setup stands out from Intel and Arm

AMD isn't the only one benefiting from the AI CPU boom.

Intel (INTC), Arm Holdings (ARM), Qualcomm (QCOM), and Nvidia are all competing for the same market share.

But BofA believes AMD has one of the clearer near-term setups.

Stronger CPU pricing would be a boon to Intel, especially if greater server pricing could offset decreased PC demand. But BofA sees Intel's value share for server CPUs falling to around 24% in 2030 from around 41% in 2025.

Arm could have the best story for long-term share gains.

BofA anticipates ARM-based merchant and custom CPUs to make substantial progress as Nvidia, Qualcomm, and hyperscale customers crank up new designs. However, Arm's royalty business continues to be affected by lower smartphone unit volumes in the near term, and the greater server opportunity is more likely to materialize in the second half of 2026 and 2027.

BofA is wary of Qualcomm's data-center ambitions due to mobile pressure, Apple risks, and execution questions.

AMD is therefore better positioned in the near term.

The company has a visible server CPU business, a GPU pipeline that's moving toward rack-scale products, and a near-term AI event where it can tie those pieces together for investors.

Leon Neal / Getty Images

Key takeaways for AMD investors

• BofA raised its AMD price target to $620 from $550.
• The new target implies about 11% upside from the AMD price listed in the report.
• The firm sees AMD benefiting from both CPU and GPU demand in AI infrastructure.
• BofA expects server CPU demand to grow sharply as agentic AI increases CPU workloads.
• AMD's Venice CPUs could strengthen its high-core-count advantage.
• The MI455X and Helios rack system are key second-half catalysts.
• The July 23 Advancing AI event could reshape how investors view AMD's AI platform.
• The main risks are PC and gaming weakness, execution on rack-scale products, and the timing of GPU ramps.

Those points sharpen the argument for stock.

AMD is still not the same as Nvidia. It doesn't have the ecosystem, the market share, or the full-stack domination that Nvidia has.

However, BofA's analysis indicates AMD might not need to dominate the AI market in the same manner Nvidia did.

AMD's opportunity could be different.

It can share in CPUs. It can boost GPUs. It can combine into rack-scale systems. And it may help if agentic AI makes CPU performance a bigger bottleneck for data-center productivity.

That provides AMD a more complete AI story than the market seems to be pricing in.

AMD's real AI test is bigger than Nvidia

The best way to describe AMD is as a competitor to Nvidia.

And that is the most useless way too.

The better investor question is whether AMD can become one of the few companies with enough CPU and GPU depth to matter in the next generation of AI infrastructure.

BofA argues it can.

The firm's high price target suggests a notion that AMD's server CPU opportunity is expanding at the same time its AI GPU path is becoming clearer. That combination gives the company more ways to succeed than investors realized in the first part of the AI boom.

There are still dangers.

And AMD has to deliver on Helios. It has to demonstrate that it can ramp MI455X with significant customers. That has to keep growing its server CPU share while Intel, Arm, and Nvidia all push harder into the same market.

But the setup is more complex than a basic GPU-share narrative.

If agentic AI makes CPUs a more vital piece of the data-center puzzle, then AMD's EPYC business is a bigger AI asset. If Helios ramps effectively, AMD can offer a broader system-level offering. And if the July event creates new momentum for customers or products, investors may need to lower their view of the company's AI ceiling.

That's the essence of the stock call.

AMD's AI edge may not be lurking in one processor.

It may be lurking in the entire system.

Related: Goldman Sachs resets AMD stock price target for the rest of 2026

The Arena Media Brands, LLC THESTREET is a registered trademark of TheStreet, Inc.

This story was originally published July 15, 2026 at 12:13 PM.