What’s new with compute: Scaling core and agentic workloads

Google Cloud announces new compute capabilities for general-purpose and agentic AI workloads with improved performance and cost-efficiency.

• •Axion N4A processors deliver 2x better price-performance than x86 VMs
• •GKE Agent Sandbox (GA) with Axion N4A offers 30% better price-performance for agent execution
• •Axion C4A.metal instances (preview) provide bare metal computing without virtualization overhead
• •C4 with Intel Xeon 6 achieves 13% better price-performance for AI inference and vector search
• •C4N and M4N/Z4D instances optimize network and memory/I/O-intensive workloads respectively

This summary was automatically generated by AI based on the original article and may not be fully accurate.