Highly Customized Optical Networking Critical for Google’s Tensor Processing Units (TPUs)

Google’s system leverages optical circuit switching (OCS) to create direct, low-latency optical paths between TPU chips, minimizing signal conversion losses. They avoid repeated optical-electrical-optical transformations. This enables efficient data sharing across thousands of chips in a single pod, with signals remaining in the optical domain for most inter-chip communication. Ironwood, seventh generation TPU, is purpose-built …

Read more

Elon, Nvidia and Saudi Will Build $25 Billion, 500MW AI Data Center

Elon Musk announced a major partnership between his AI company xAI and Saudi Arabia’s state-backed AI firm Humain (backed by the Suadi Public Investment Fund) to build a 500-megawatt (MW) AI data center in the kingdom. It will use Nvidia GPUs and will be xAI’s largest and first data center project outside the United States. …

Read more

Google Project Suncatcher to Put TPUs for AI in Space in 2027

Google Project Suncatcher is a new research moonshot to one day scale machine learning in space. Working backward from this potential future, they are exploring how an interconnected network of solar-powered satellites, equipped with our Tensor Processing Unit (TPU) AI chips, could harness the full power of the Sun. The next step is a learning …

Read more

Nvidia H100 is in Space on Starcloud-1

The Starcloud-1 satellite launched this Sunday and has successfully separated from the SpaceX rocket that brought it to orbit. About the size of a small refrigerator, the 60kg satellite features the first Nvidia H100 GPU in space. StarCloud CEO Philip Johnston said Starcloud was founded only 21 months ago and this satellite hosts the first …

Read more

Global AI Data Centers Keep Growing Despite Adjustments From Amazon and Microsoft

AWS and Microsoft are pausing or slowing down the development of data centers, according to a Wells Fargo research note. Several industry sources told Wells Fargo’s analysts team that AWS has “paused a portion of its leading discussions” on colocated data centers. Microsoft capex for the quarter came in at $21.4bn this was $1.2 billion …

Read more

META Custom AI Inference Chips MTIA for a META AI Personal Assistant

META CEO Mark Zuckerberg predicts that 2025 is the year that an AI assistant will serve over one billion users and he thinks META will be company that provides that AI assistant. In 2023, Meta revealed they have AI inference accelerators that they are designing in-house specifically for Meta’s AI workloads. Deep learning recommendation models …

Read more

2025 The OMG Year of AI and Humanoid Robots

The billionaire investors and entrepreneurs at the All in Podcast described 2025 as the year of AI and robots. I, Brian Wang, and Randy Kirk reviewed why and how AI and robots will be transforming the world. Summarizing the key points and claims and providing context. Dave Frieberg says 2025 is the year when people …

Read more