Nvidia CEO Jensen Huang predicts that data center operators will spend $1 trillion upgrading their infrastructure over the next five years to support demand from artificial intelligence (AI) developers. A significant portion of this spending will be on graphics processing units (GPUs), which are ideal for AI training and inference.
While Nvidia is currently the top supplier of data center GPUs, Advanced Micro Devices (AMD) is gaining ground with its new MI300X GPU. AMD’s flagship H100 GPU was released in 2022, but uptake has been promising in 2024, with many of Nvidia’s top customers adopting the MI300X for better performance and cost savings.
However, AMD is not stopping there. The company plans to release its MI325X this quarter, which boasts up to 20% better AI inference performance than Nvidia’s H200. Additionally, AMD’s new Ryzen AI series of central processors (CPUs), neural processors (NPUs), and GPUs are designed to process AI workloads on-device, reducing reliance on external data centers for a faster user experience.
Recently, AMD reported record revenue of $6.8 billion in Q3 2024, driven by strong GPU sales in its data center segment. The company’s client segment, which includes the Ryzen AI series, grew revenue by 29% to $1.9 billion. With its innovative pipeline, including the upcoming MI350 series, AMD is poised to challenge Nvidia’s dominance in the AI chip market.
Key points to consider:
* Nvidia’s market share will decline over time as competition grows
* Buying AMD stock could be a hedge for existing Nvidia investors
* AMD’s revenue growth and innovation pipeline make it an attractive long-term investment opportunity
Source: https://www.fool.com/investing/2024/11/04/nvidia-stock-up-173-in-2024-stock-buy-wall-street