Amazon Invests an Additional $5 Billion in Anthropic! The AI Battlefield Enters the “In-House Chip” Era with a $100 Billion Cloud Contract

📰 News Summary

• Amazon has made an additional investment of $5 billion in Anthropic, raising the total investment to $13 billion.
• Anthropic has committed to spending over $100 billion on AWS cloud services over the next decade.
• They have secured new computing capacity of up to 5GW (gigawatts) for training and executing Claude.

💡 Key Points

• The contract includes the use of Amazon’s proprietary AI accelerator chips, from the upcoming “Trainium2” to the not-yet-released “Trainium4.”
• At the heart of this partnership is the utilization of the low-power CPU “Graviton” and the “Trainium” chips, which compete with NVIDIA.
• Similar to the $5 billion investment in OpenAI two months ago (when it was valued at $730 billion), this strategic partnership involves not just cash but also infrastructure support.

🦈 Shark’s Eye (Curator’s Perspective)

It’s incredibly exciting that Anthropic is banking its future on Amazon’s proprietary “Trainium” series! The commitment to future supply of the yet-to-be-released “Trainium4” illustrates just how fierce the competition for computing resources has become. The scale of “5GW” secured in this deal is practically on a national infrastructure level! This contract proves that the era dominated by NVIDIA alone is over, as cloud giants with in-house chips and top-tier AI developers join forces in a “vertical integration model.”

🚀 What’s Next?

Anthropic is likely to soon announce a new funding round valuing the company at over $800 billion. The next-generation Claude, fully utilizing Amazon’s “fangs (chips),” is expected to dominate the market with efficiency and cost performance!

💬 Haru-Shark’s Insight

Shark reporter “Haru-Same”: A $100 billion shopping spree is too massive even for a shark’s stomach to digest! We can’t take our eyes off this powerful symbiotic relationship between Amazon and Anthropic! 🦈🔥

📚 Terminology

• Trainium: A proprietary accelerator chip designed by Amazon (AWS) for AI learning and inference, known for its high cost-performance compared to NVIDIA GPUs.

• Graviton: An ARM-based custom CPU developed by AWS, boasting high processing power while maintaining low power consumption, supporting the backend of AI systems.

• GW (Gigawatt): A unit of electrical capacity, used in this context as an indicator of the scale of computing resources in data centers running AI models.

• Source: Anthropic takes $5B from Amazon and pledges $100B in cloud spending in return