Forbes·2025-11-28 13:55Core Viewpoint - The narrative that AMD is the next Nvidia is misleading, as current financial metrics suggest AMD is overvalued compared to Nvidia [1][4]. Valuation Comparison - Nvidia trades at 38 times consensus earnings with a revenue growth rate exceeding 65%, while AMD trades at 55 times consensus 2025 earnings with a revenue growth rate of 32% [4]. - AMD's pricing is compared to a "champagne price for a beer budget," indicating that investors are paying a premium for AMD without corresponding performance [4]. Demand and Supply Analysis - The argument that AMD suffers from supply constraints is debunked; AMD's lower pricing indicates a demand issue rather than a supply problem [7]. - AMD's profit margins are significantly lower than Nvidia's, with AMD retaining about 24 cents per dollar compared to Nvidia's 65 cents, reflecting their different market positions [8]. Strategic Positioning - AMD should be viewed as a cost-effective alternative in the AI market rather than a direct competitor to Nvidia [9]. - The shift from AI training to inference is highlighted, where cost efficiency becomes crucial, making AMD's offerings more attractive for companies looking to reduce expenses [11][16]. Memory Capacity Advantage - AMD's MI300X chip offers a significant memory advantage, allowing companies to run large AI models on a single chip, which can lower total costs [10][13]. - This memory capacity is particularly relevant as companies transition from high-cost training to more cost-effective inference operations [11]. Market Dynamics - Big Tech companies are incentivizing AMD's survival to maintain competitive pressure on Nvidia, ensuring that they have an alternative supplier to mitigate Nvidia's pricing power [15]. - The need for a duopoly in the market is emphasized, as companies are not just investing in AMD's technology but also in the necessity of having multiple suppliers [15]. Conclusion - The recommendation is to invest in AMD not for its potential to outperform Nvidia but for its role as a cost-effective solution in the evolving AI landscape [16].