A $5.25 million Series-A investment has been raised by the Indian decentralized cloud computing initiative GPU.NET. As more people are experimenting with novel use cases for artificial intelligence (AI) and machine learning (ML) techniques, GPUnet seeks to operate reasonably priced pre-training clusters for Enterprise customers to rent GPU compute - a highly precious resource in the present environment.
The world's four biggest cloud providers - AWS, Microsoft Azure, Google Cloud, and Oracle - would control roughly 80% of the parallel computing resources worldwide, forcing funded startups and researchers to either purchase it from them at a steep cost or obtain their own GPUs, which require experienced data center management in addition to lengthy wait times for NVIDIA chips to arrive.
Thus, where is the supply of GPUnet sourced from? It originates from independent data center operators that have been lending their computing power to Web3 mining initiatives. Over time, they have acquired the expertise required to manage data centers and provide computing power for cloud-accelerated and HPC applications. Significant computing capabilities would often be accessible from these suppliers, but only in small clusters. To unify them on a single network, the GPUNet team intends to use a distributed computing architecture, which would aid in offering developers and researchers an intuitive cloud environment.
Momentum6, Spicy Capital, Exnetwork, Blackdragon, Zephyrus Capital, Aza Ventures, F7 Foundation, Halvings Capital, and Bigger than Race Ventures are participating in GPU.NET's $5.25 million Series-A fundraising round.
Public Blockchain
With the public blockchain that GPUnet has created, any providers - big or small - can participate on an even playing field. They are developing state-of-the-art applications like federated learning and fully homomorphic encryption in addition to creating an environment that is favorable for AI and gaming applications.
“In 2030, I see a world where energy and computation alone will generate a trillion-dollar GDP. Significant advancements in the AI application sector will make this possible,” said the founder and CEO of GPU.NET, Suraj Chawla. “This is achievable if the ownership of the processing resources is decentralized. If the largest AI firms today are controlled by the corporations that possess the largest data centers, like Microsoft and Meta, this will result in a significant monopoly or duopoly in the LLMs that are consumed on a worldwide scale.”
A public blockchain, often termed as a permissionless blockchain, is an open network accessible to anyone for participating and viewing transactions. It's fundamentally decentralized with no single authority in control, embodying a distributed ledger that records transactions transparently. Some crypto examples include Bitcoin and Ethereum. These blockchains allow anyone to engage in activities like mining and transaction validation, fostering a trustless environment where transactions are executed peer-to-peer without intermediaries. They are also resistant to censorship, ensuring no single entity can manipulate or restrict participation. Contrastingly, private blockchains restrict access to certain approved users and are managed by specific individuals or organizations, offering scalability and confidentiality at the cost of open access. Public blockchains ensure transaction data immutability and are maintained by participants who are incentivized with digital currency, enhancing network security and integrity.
