AMD's Strategic AI Chip Investment: If You Can't Beat Them, Buy Them

When it comes to AI chips, Nvidia is the dominant player. However, AMD is rapidly catching up. Recently, AMD made a significant move by investing $655 million in the Finnish AI startup Silo AI, one of Europe's largest private AI labs.

This investment gives AMD access to Silo AI's 300-person team, experienced in building custom large language models (LLMs), which power products like ChatGPT. This is part of AMD's broader investment strategy, with over $125 million invested in AI-related acquisitions in the past year, including companies like Mipsology and Nod.ai.

Key Points

• AMD announced a $655 million investment in Finnish AI startup Silo AI.
• The investment aims to strengthen AMD's enterprise AI product ecosystem.
• AMD's chips are becoming a key competitor in the AI market, despite Nvidia's lead.

Why The Silo AI Investment Makes Sense for AMD

This investment allows AMD to move beyond being just a chipmaker. By partnering with Silo AI, AMD gains access to top AI engineers with extensive experience in developing AI models. According to Vamsi Bopping, senior vice president of AMD's Artificial Intelligence Group, this partnership will accelerate AMD's AI strategy and help develop and implement AI solutions for global customers.

In 2023, Silo AI generated €21 million in revenue, adding clients like Philips, Rolls-Royce, Nvidia, and Intel. This success suggests that AMD will be able to enhance its product offerings with compelling AI-powered solutions.

AMD's Position in the AI Chip Market

The AI chips market is expected to grow dramatically, from $16.86 billion in 2022 to $227.48 billion by 2032, according to Precedence Research. AMD holds the second-largest market share behind Nvidia, which dominates with 70-95% of the market. Despite this, AMD is doing well, with CEO Lisa Su expecting AI chip sales of $4 billion by the end of 2024. However, Nvidia's strong market position remains a significant challenge for AMD's long-term growth.

Currently, AMD's CPUs are popular, but many organizations prefer Nvidia GPUs for intensive AI workloads.

AMD vs. Nvidia in the AI Chip Race

In the AI chip market, the competition is between AMD's Radeon GPUs and Nvidia's H100 and H200 GPUs. The key difference is performance and cost-efficiency. Olivier Blanchard, Research Director at The Futurum Group, notes that Nvidia's GPUs are currently the industry standard for high-end AI workloads. However, AMD's new MI300X GPUs are excellent alternatives and sometimes even outperform Nvidia's.

Blanchard predicts that AMD will continue gaining momentum in the data center GPU segment through 2025, despite Nvidia's market penetration advantage. In the PC GPU segment, AMD's Radeon GPUs compete well with Nvidia's RTX GPUs, offering competitive performance and price advantages.

The Bottom Line

AI chips are crucial for the development and deployment of AI models. AMD has built a robust product ecosystem, positioning itself as a viable competitor to Nvidia. While Nvidia is unlikely to lose its market share soon, AMD provides a necessary alternative for enterprise AI development investments.