Which AI Stock, NVIDIA or Broadcom, Is the Better Bargain to Buy?
Advancements in artificial intelligence (AI) have made NVIDIA Corporation NVDA a Wall Street darling for some time. On the other hand, Broadcom Inc.’s AVGO recent share repurchase plan, along with its planned moves to capitalize on the growing AI field, has delighted income-oriented investors.
However, the latest trade war between the United States and several other countries has caused market turmoil, with shares of NVIDIA and Broadcom declining 19.9% and 25.7% year to date. This presents an opportunity to buy these quality stocks at a discounted price for long-term gains. But which stock is the better bargain? Let’s explore –
Image Source: Zacks Investment Research
NVIDIA leads the growing graphics processing unit (GPU) market with over 80% share, giving the Jensen Huang-led company a competitive edge over its peers. More than Advanced Micro Devices, Inc.’s AMD ROCm software platform, NVIDIA’s CUDA software platform is in demand among developers.
At the same time, there is an insane demand for NVIDIA’s new Blackwell chips among notable tech firms due to their energy efficiency and faster AI interfaces. The older Hopper chips also maintain steady demand due to their higher quality compared to rival Intel Corporation’s INTC chips (read more: Can Intel Be the Turnaround Stock of 2025? Is the Time to Buy Now?).
NVIDIA, meanwhile, is well-poised to make the most of the increase in AI data center spending. Top cloud computing stocks are fulfilling the increase in demand for AI workloads by acquiring GPUs and are spending $250 billion on AI data center infrastructure. One of NVIDIA’s prominent customers, Microsoft Corporation MSFT, may have reduced data center spending, but Alphabet Inc. GOOGL and Amazon.com, Inc. AMZN are now supporting NVIDIA in this aspect.
Broadcom expects the demand for its custom AI accelerator to increase in the near future and the market size for its execution processing units (XPUs) to touch $60-90 billion by fiscal 2027, way higher than last year’s $12.2 billion.
Broadcom’s XPUs can work on a particular workload and even outdo NVIDIA’s GPUs. Moreover, they have the competence to perform tasks more affordably than GPUs. Thus, banking on the rise in demand for accelerator chips, Broadcom is now developing a 2-nanometer AI XPU, which is unique. The company’s application-specific integrated chips (ASICs), known to support AI and machine learning, are witnessing strong demand as well.
No doubt, with GPUs and XPUs gaining popularity in the coming years, and AI infrastructure spending increasing, the future does look bright for both NVIDIA and Broadcom. However, from a long-term perspective, NVIDIA is a better choice than Broadcom.