Can the struggling AMD become a beneficiary of the "Deep Sea Shock"?

According to MOOMOO, AMD's sales revenue for Q4 of fiscal year 2024 increased by 22.01% to $7.526 billion compared to the same period last year.According to MOOMOO, AMD's revenue for the fourth quarter of 2024 increased by 22.01% year-on-year to $7.526 billion.、It is expected that EPS will increase by 46.78% to $0.60 compared to the same period of the previous year.。

In the third quarter of 2024, AMD achieved an 18% revenue growth year-on-year, reaching $6.8 billion (surpassing market expectations of $6.7 billion) due to strong performance in the Datacenter business. Non-GAAP operating margin increased by 300 basis points to 25%. Datacenter revenue increased by 25% from the previous quarter and by 122% year-on-year, demonstrating continuous growth in AMD's Datacenter market.

Stock prices are declining, with decreased value.
In 2024, while AI-related stocks soared, AMD's stock price struggled. AMD's stock price fell by 18% in 2024, significantly dropping from its 52-week high of $227 recorded in March 2024, and is currently trading around $117. AMD's estimated EPS for fiscal year 2025 is $4.99. The FWD price-to-earnings ratio (P/E) is 24.03. Additionally, the FWDP/E for fiscal year 2026 is expected to decrease to 17.3 times. The forecasted PEG ratio is 0.82, significantly below the industry average of 1.89. The current valuation reflects the market's cautious stance on AMD's growth outlook.

AMD is not NVIDIA or Broadcom.
In the field of AI chips, AMD released the MI325X chip in October 2024, competing with NVIDIA's Blackwell H200 Tensor Core GPU. While AMD claims superiority in aspects like memory capacity and bandwidth over NVIDIA's H200, NVIDIA's CUDA programming language is entrenched among AI developers, positioning NVIDIA advantageously in the market while maintaining market share.

Furthermore, AMD faces competition from other technology giants like Apple, Amazon, and Google. These high-tech companies are developing AI-specific integrated circuits (ASICs) with Broadcom. Amazon has developed the Graviton chip, achieving a 40% efficiency improvement in AI model inference and development. In addition, Microsoft introduced two new AI infrastructure chips in November 2024 to enhance AI computation speed and data security. AMD faces increased pressure on its market share in the AI field.

Highlights of CES 2025
At CES 2025, AMD announced updates to its AI processor lineup and next-generation graphic technology. It includes the Radeon RX 9000 graphics card featuring RDNA 4 technology, excelling in AI tasks, multimedia processing, and ray tracing. Additionally, AMD unveiled multiple AI-capable processor series like Ryzen AI Max, Ryzen AI 300, and Ryzen 200. AMD also introduced the high-end thin and light notebook processor series called the "Ryzen AI Max series". This series is designed to strengthen next-generation AI applications with up to 16 fully functional CPU cores based on Zen 5 architecture, a maximum of 40 GPU compute units, and over 50 TOPS AI processing power.

AMD has integrated AI capabilities into all of its processors and graphics solutions, adopting a broader strategy.

Will DeepSeek bring new opportunities?
In the short term high-tech stock market, investors are concerned that DeepSeek's leadership in the 'low-cost AI large model computing power paradigm' may lead to a significant reduction in orders for AI GPUs from technology giants. Additionally, the wave of low-cost computing power led by DeepSeek has raised questions among investors about the reasonableness of spending by major AI companies in the USA.

However, as low-cost computing power AI large models accelerate entry into multiple application fields such as enterprise software, household appliances, PCs, robotics, etc., the demand for AI computing power is expected to continue to grow.

On January 7th, AMD and DeepSeek announced that the DeepSeek-V3 model will be operated on AMD Instinct™ GPUs, optimized for supporting SGLang. With AMD's ROCm™ software and Instinct™ GPU accelerator, DeepSeek-V3 will be supported at various important stages of development. This partnership strengthens AMD's collaboration with the open-source community and is expected to enhance its influence among AI developers. The utilization of FP8 technology and ROCm platform could help AMD compete for more market share and provide strong support as it competes with rivals like NVIDIA.

While AMD faces market competition pressure and a decline in stock prices, its business expansion in the AI chip sector, technological innovation, and the reduction of AI model costs by DeepSeek could potentially bring about performance growth in products for small and medium enterprises as well as individual consumers.