Singapore – Oct 29, 2024 – Aethir, a provider of decentralized cloud infrastructure, has partnered with GAIB, the economic layer for AI and compute, and GMI Cloud, an AI and AGI-focused cloud service provider to integrate H200 Tensor Core GPUs into its decentralized computing ecosystem.
The companies said this marks the first production-ready deployment of H200 GPUs in the Web3 space, providing unprecedented access to high-performance computing technology for enterprises and developers. H200 Tensor Core GPUs, the latest high-performance graphics processing unit (GPU), are engineered to accelerate AI and machine learning workloads.
Built on the advanced Hopper architecture, the H200 delivers significant advancements in memory capacity, bandwidth, and computational efficiency compared to its predecessor, the H100. These enhancements make the H200 the ideal choice for enterprises and developers tackling complex deep learning and generative AI applications. Additionally, this collaboration will also make GPU resources more accessible and cost effective on a global scale.
Through this initiative, Aethir, GAIB, and GMI Cloud will equip enterprise clients with the most powerful computing power required to handle demanding GPU tasks, including large language model (LLM) training and AI inference. By introducing a decentralized approach to accessing the latest GPU technology, this integration accelerates the development of large-scale AI applications by providing companies with a new, decentralized way to access the latest GPU technology.
GAIB introduces a new type of yield-bearing asset backed by real AI cash flow. GMI Cloud stakes physical H200 GPUs on the GAIB network, which are then re-staked to Aethir, making these compute resources available to anyone who needs them. In the future, GAIB will tokenize these GPUs and their yields, bringing them on-chain. This will enable users to directly invest, trade, and own GPUs as assets—an unprecedented opportunity. By purchasing GPU NFTs, users can earn rewards and yields generated by the underlying GPUs that power both Web2 and Web3 AI demands.
As a key partner in this collaboration, GMI Cloud also plays a crucial role in optimizing the integration of H200 GPUs into the decentralized network, leveraging its cloud expertise and technical resources. GMI’s extensive experience in cloud infrastructure ensures the full potential of H200 GPUs is realized, driving innovation in AI development and enhancing accessibility for developers worldwide.
Alex Yeh, Founder and CEO of GMI Cloud, said “our mission is to empower humanity’s Al ambitions with an efficient, on-demand GPU cloud,” said Alex Yeh, Founder and CEO of GMI Cloud “We’re not just building a cloud, we’re creating the backbone of the Al era. By joining forces with two powerful industry players, GMI Cloud is transforming how developers and data scientists utilize NVIDIA GPUs, driving Al innovation for the benefit of all.”
“Through this integration, we’re delivering enterprise-grade decentralized GPU computing on a global scale,” said Daniel Wang, CEO of Aethir. “By leveraging our vast network, we’re empowering the next generation of Al developers with the tools they need to efficiently build, train, and deploy powerful models.”
“GAIB is solving the challenges of investing in illiquid compute assets and high barriers to entry by building an economic layer that turns GPUs into liquid, tradable yield-bearing assets,” said Kony, CEO of GAIB. This approach unlocks new investment opportunities, enhances market efficiency, and accelerates the growth of the Al economy.”
