Why AMD stock slipped as much as 5% today

Shares of Advanced Micro Devices declined sharply on Tuesday following reports that the company is facing manufacturing challenges with its latest artificial intelligence accelerators.

The reports raise concerns about delays to its data-centre expansion plans.

AMD stock was down about 2.5% in midday trading, recovering slightly after falling as much as 5% earlier in the session.

The pullback followed reports that production of the company’s Instinct MI455X accelerators has been delayed as engineers work to resolve technical issues.

According to the reports, the production slowdown is also expected to affect shipments of AMD’s Helios rack-scale systems, limiting how many units the company can deliver this year.

As a result, the market is said to be expecting only low-volume shipments of the affected products in 2026.

Manufacturing setbacks hit AI ambitions

The reported delays represent a setback for AMD’s efforts to expand its footprint in the rapidly growing market for AI accelerators.

The Instinct MI455X is part of the company’s broader push to compete more directly with Nvidia in high-performance computing and generative artificial intelligence workloads.

Any disruption to production risks slowing AMD’s ability to scale its data-centre offerings at a time when demand from cloud providers and AI developers remains strong.

Delays could also weigh on near-term revenue growth tied to large-system deployments.

The reports suggest that the manufacturing issues must be resolved before mass production can begin, pushing meaningful volumes further into the future.

Analyst on AMD stock

DA Davidson analyst Gil Luria said AMD’s Instinct products look “great on paper,” but argued that customers are focused on real-world performance at scale rather than headline specifications.

He said that while AMD can point to competitive specifications for chips such as the MI300X, MI350X, and MI400X, large-scale deployments depend heavily on interconnects and systems integration — areas where Nvidia holds an advantage and AMD “doesn’t have a serious answer.”

According to Luria, handling frontier workloads across tens of thousands of GPUs is fundamentally a systems-level challenge.

As a result, he said, the cost of achieving meaningful performance with AMD’s solutions may be higher than advertised.

Luria also said AMD is increasingly playing “catch-up” in the AI accelerator market, reflecting its slower progress in building a fully integrated platform.

Luria recently initiated coverage of AMD with a neutral rating and a $220 price target, reflecting a cautious stance on the company’s competitive positioning in AI infrastructure.

He also initiated coverage of Taiwan Semiconductor Manufacturing Company and Intel, assigning TSMC a Buy rating, while giving neutral ratings to AMD and Intel.

Broader ecosystem gaps

Beyond hardware performance, Luria highlighted weaknesses in AMD’s broader ecosystem.

He said high-performance computing depends not only on GPUs, but also on software, networking, memory, and system-level integration.

In his view, AMD is “meaningfully lacking outside their core hardware capabilities.”

By contrast, competitors such as Nvidia and Broadcom offer more comprehensive portfolios, enabling faster time-to-deployment for customers.

Luria said AMD remains more dependent on third-party networking vendors, which can complicate large-scale rollouts.