exaBITS unlocks millions of consumer GPUs for AI computing.

exaBITS launched the world's first truly accelerated decentralized computer network. By pooling global GPU resources and optimizing their use, they can significantly reduce training costs and deploy large AI models.

Already, exaBITS has deployed more than 65,000 GPUs globally in its private beta phase, attracting customers from startups to well-known unicorns in the AI ​​industry and saving hundreds of thousands of dollars. The company's unique supply reach and growing consumer-grade GPUs have resulted in an ever-growing active customer base.

This foundation of unparalleled cost efficiency, accelerated execution and access to the machines necessary for their projects, positions exaBITS as the leading choice in the global AI sector.

One example is with their large enterprise customers who used an institutional cloud computing provider early in their journey. They quickly realized that the high operating costs of this calculation were a huge drain on resources at over $150k per month. When they switched to exaBITS, their computing costs decreased by 71 percent and their performance increased by 130 percent.

The challenge in AI is finding powerful GPUs like Nvidia's H100, A100, and AMD's MI300 and MI200, which are highly demanding and not very efficient as a result. exaBITS addresses not only these high-end GPUs, but also consumer-level GPUs such as the 3090/4090, modifying them to match their performance.

This capability is critical considering that many companies face years-long waiting lists to get enterprise-class GPUs or bottlenecks to secure instances at other cloud computing providers.

H100/A100 supply and consumer GPUs

At the heart of the exaBITS network are T3 data centers, mining facilities and massive distributed consumer-grade GPUs, consisting of more than 65,000 GPUs, including many H100s and A100s. The key to delivering AI at scale at an affordable price is the ability to boost performance of weak GPUs via blockchain.

They achieved up to 15x cost reduction and 90% lower costs, paving the way for a robust decentralized network that efficiently utilizes consumer GPUs. This reduces the dependence on enterprise-level machines, increasing the availability of resources in the network.

The future roadmap

exaBITS has proven its ability to comprehensively meet the technical needs of the market and its roadmap shows a clear path to the future. Part of the roadmap includes a ML DevOps Platform that intelligently manages and accelerates training and awareness, AI models for product efficiency and service quality.

This not only dwarfs exaBITS, but also surpasses incumbents like AWS and Google. In addition to acceleration, exaBITS deploys its anti-destruction technology. This solves the hidden killer of AI projects, the massive failure of GPU clusters. This technology improves system stability by more than 50%, which greatly increases the efficiency of resource use.

Their researchers have built an automatic recovery feature that instantly detects the moment the GPU crashes on a live stream and allows months of training to restart and continue where training failed, reducing the risk of long-term training and increasing costs during workloads. It has been discontinued.

With a rich background in companies such as Bitmain, Meta and (from data centers or distributed computing) and various mobile operators, the team has extensive experience in brokering deals for data centers and mining facilities.

The exaBITS team consists of scientists with deep expertise in cloud computing, AI and blockchain, supported by a business team well-versed in Web3 business models. Together, they bring more than 50 years of combined experience in these fields.

exaBITS is exploring underutilized GPUs in a technology stack and team that can use them for powerful AI insight and training. With access to high-end GPUs like the H100 and A100 and the ability to make consumer GPUs work like enterprise-grade GPUs, exaBITS positions itself as a premier decentralized cloud computing provider, democratizing AI development and lowering cost barriers for AI projects and developers globally.

Visit exaBITS and learn more