Tag Archives: custom hardware
Google Frozen v2 Chip: Revolutionizing AI Efficiency with Gemini Silicon
Discover how the new Google Frozen v2 chip aims to hardwire Gemini architecture into silicon, achieving 10x greater efficiency for future AI inference. Continue reading
Google Marvell AI chips: Advanced Talks for Custom AI Processor Development
Google and Marvell are reportedly negotiating a deal to co-develop Google Marvell AI chips, including new memory and tensor processing units to boost cloud infrastructure efficiency. Continue reading