Stanford lab, focused on AI research, will use the decentralized cloud computing platform Theta EdgeCloud for its work on LLMs.
The decentralized cloud could be the solution for AI’s vast computing needs. On April 17, Theta Labs revealed that Stanford University’s AI would use Theta (THETA) EdgeCloud in its work on large language models. The lab, led by Assistant Professor Ellen Vitercik, will use the platform for discrete optimization and algorithmic reasoning of LLMs.
Stanford joins a growing list of academic institutions using the decentralized platform for research. According to Theta Labs, other adopters of EdgeCloud include Seoul National University, Korea University, the University of Oregon, Michigan State University, and more.
Big tech and decentralized services compete for AI compute
Big tech companies have been rapidly expanding their investment in computing infrastructure, especially one geared toward powering AI. In 2024, Microsoft invested $3.3 billion in a data center in Wisconsin, with the backing of Joe Biden’s administration.
At the same time, Amazon stated it plans to spend $11 billion on data centers in Indiana. Google, on the other hand, is going global, investing $1.1 billion in its data center in Finland and building another one in Malaysia for $2 billion.
Still, the big tech model isn’t the only one competing for AI workloads. Unlike most traditional LLM services, Theta EdgeCloud operates as a decentralized cloud computing platform. Its infrastructure is geographically distributed, meaning it does not rely on massive centralized data centers to provide compute power.
Instead, the platform uses blockchain technology to reward smaller GPU providers based on the revenue they generate from users. This allows Theta to operate with lower capital expenditures and scale more quickly. In turn, it offers more affordable infrastructure for users.
Theta Network is a blockchain protocol originally designed for decentralized video streaming. However, the network has since expanded to provide decentralized infrastructure for cloud computing, with a particular focus on AI applications.
