Source URL: https://www.theregister.com/2025/02/05/amd_q4_fy2024/
Source: The Register
Title: Who’s afraid of DeepSeek’s impact on AI hardware sales? Not AMD CEO Lisa Su
Feedly Summary: Predicts more efficient ML architectures will drive adoption, see Instinct sales accelerate, shares dive
AMD’s chief exec Lisa Su has predicted the chip designer’s Instinct accelerators will drive tens of billions of dollars in annual revenue in coming years, despite DeepSeek-inspired speculation that next-gen AI models may not need the same level of compute infrastructure used to produce such tools today.…
AI Summary and Description: Yes
**Summary:** The text provides insights into AMD’s financial performance and future prospects, particularly focusing on its Instinct accelerators vital for AI workloads. Lisa Su highlights the impact of efficient compute models on AI adoption, while addressing competitive threats from AI ASICs. This information is crucial for professionals in AI, cloud computing, and infrastructure security as it reveals trends affecting hardware investments in AI capabilities.
**Detailed Description:**
The analysis of AMD’s Q4 earnings call highlights key developments in the company’s strategy regarding AI compute infrastructure, revealing significant implications for the future of AI and its associated hardware. Key points include:
– **Growth Prediction for AI Accelerators**:
– AMD CEO Lisa Su projected that revenues from Instinct accelerators would increase significantly beyond the existing $5 billion mark expected in FY 2024.
– The anticipated strong demand for AMD’s next-gen MI355X accelerators is expected to drive revenue growth in the second half of FY 2025.
– **Emergence of Efficient AI Models**:
– Su referenced DeepSeek’s LLMs that require less compute infrastructure, raising questions about the historical investment in extensive AI infrastructures.
– The efficiency of new models potentially may suggest that organizations could optimize their spending, influencing the dynamics of the hardware market.
– **Innovations in Rack-Scale Solutions**:
– Future products will utilize the CDNA 4 architecture, enabling high-performance rack-scale architectures that integrate CPU and GPU, which is essential to meet demands at data center scale.
– **Market Environment and Competitive Insights**:
– AMD is facing growing competition from custom AI ASICs developed by companies like Google and Amazon. However, Su believes that the diversity of AI workloads will sustain demand for GPUs alongside ASICs and FPGAs.
– There is a substantial total addressable market (TAM) related to AI compute solutions, projected to exceed $500 billion.
– **Strong Performance and Strategic Outlook**:
– AMD’s datacenter division showed significant growth, with revenues jumping to $3.9 billion in the fourth quarter, evidencing the central role of Instinct GPUs in AMD’s success.
– Despite challenges in AMD’s gaming and embedded divisions, the company remains optimistic about recovery in those segments.
This information underlines the ongoing evolution of AI hardware requirements and sets the stage for future developments in AI infrastructure, which will directly impact compliance, security posture, and operational strategy within organizations aiming to leverage AI capabilities. Understanding AMD’s strategic direction, market trends, and technological advancements is crucial for security and compliance professionals as they navigate the complexities of AI adoption and infrastructure investments.