AMD or Nvidia: Load Up on One AI Chip Stock, but Avoid the Other, Says Investor

Advanced Micro Devices (NASDAQ:AMD) might be a chip giant boasting a market cap of around $772 billion, but the company continues to operate in the shadow of Nvidia (NASDAQ:NVDA). Despite AMD’s rapid growth and strengthening position in the AI chip market, it is still often viewed as a smaller version of its larger peer.

Summer Sale - Claim 70% Off TipRanks

Explore NVDS for 2X short leverage on NVDA

However, that does not mean this is how things will remain. In fact, investor Elizabeth Pramila believes the Lisa Su-led company is poised to outperform its bigger rival over the coming years. The core of her thesis is that AMD does not need to surpass Nvidia in raw GPU performance to win. Instead, it is positioning itself to capture a growing share of AI spending by competing on efficiency, flexibility, customization and the economics of running AI workloads.

The biggest opportunity could come from the shift toward inference. While Nvidia has built an enormous business around training ever more sophisticated AI models, inference occurs continuously whenever those models are used. As AI becomes embedded across products and services, inference could become a massive, recurring source of computing demand. Pramila believes AMD’s focus on delivering more inference tokens per dollar could therefore become a meaningful competitive advantage, particularly as hyperscalers and enterprises become more concerned with the cost and return on every dollar of AI infrastructure spending.

AMD’s MI400 GPU family and Helios rack-scale platform are central to this strategy. Helios combines GPUs, CPUs, networking, memory, power delivery and cooling into an integrated AI infrastructure solution, giving customers an alternative to Nvidia’s ecosystem. The fact that major AI labs and cloud providers, including existing Nvidia customers, are adopting Helios also points to AMD making inroads into Nvidia’s established customer base.

The investor also sees AMD’s sixth-generation EPYC server CPUs as an important part of the long-term opportunity. The lineup is designed for a range of workloads, including agentic AI, inference, sovereign AI, analytics and other compute-intensive applications.

Pramila sees another advantage in AMD’s emphasis on an open and customizable software ecosystem. The investor believes this could become more important as customers seek to optimize infrastructure for specific workloads rather than simply purchase the most powerful accelerator available.

Ultimately, Pramila believes AMD has a potentially longer growth runway than Nvidia. “How I see this, to reiterate my long-standing position, is NVDA is the hare outpacing AMD the tortoise on visible metrics, but the underlying currents seem to suggest that the tortoise may be the eventual winner in this largely uneven race, where advantages arise not from sheer scale alone but from foundational efforts to set up a more long-lived revenue profile, which AMD certainly has,” the investor summed up.

To this end, Pramila rates AMD stock a Strong Buy, while her rating for Nvidia remains a Strong Sell. (To watch Pramila’s track record, click here)

Wall Street agrees with only half of that thesis, given it sees no reason to choose between the pair, as both boast Strong Buy consensus ratings. AMD’s $647.42 average price target points toward 12-month returns of 37%, while NVDA is set to gain 40.5%, considering its average target clocks in at $301.82. (See AMD stock forecast or NVDA stock forecast)