TechFlow Logo
Login/ Sign up
ETH Gas
Gwei
Fear
gas
Coinbase-Backed B3 Labs Launches B3IQ: Own Dedicated GPU Servers via Installments, Offset Purchase Costs by Renting Out Idle Computing Power

Coinbase-Backed B3 Labs Launches B3IQ: Own Dedicated GPU Servers via Installments, Offset Purchase Costs by Renting Out Idle Computing Power

2026.08.12
Share
Research teams from Stanford, NYU, and Dartmouth are already using it to run cancer research, specialized model training, and more.

Privacy is becoming an increasingly prominent issue in the AI field. Every prompt sent to a centralized AI service provider is like a deposit stored in someone else's vault: research data, business logic, and creativity flow into systems the sender cannot control, with terms not set by themselves. The most direct solution is to run AI models on owned hardware, but previously this method was out of reach for most people: high-end GPU systems cost tens of thousands of dollars, supply is tight, and operation requires space and professional maintenance.

Today, B3 Labs launched B3IQ to fill this gap: B3IQ is a pioneering AI infrastructure service designed to give universities, enterprises, and professional users greater control over the hardware, models, and data their AI workloads rely on.

Before this, institutions needing AI compute power had only two imperfect choices: one is to rent from cloud providers, meaning prices fluctuate, there is no ownership, and supply is limited when GPUs are scarce; the other is to buy the entire machine outright, requiring bearing all purchase costs at once, plus electricity, cooling, and maintenance burdens. B3IQ offers a third model combining both: Economically it is "ownership": assets and earnings belong to you; operationally it is "managed hosting": the data center and maintenance are handed over to the platform.

B3IQ users can purchase exclusive NVIDIA GPU systems via installment payments: manufactured in the US by Andromeda, an American AI system manufacturer invested by B3 Labs, and hosted in Oregon. Through the B3IQ dashboard, owners can match idle compute power with compute demand to monetize, where earnings can be used to offset the remaining purchase balance or kept as income. After paying off the full amount, owners can choose to continue hosting with B3IQ or arrange for physical hardware delivery.

Early users of B3IQ include faculty, AI researchers, and student teams from New York University, Dartmouth College, the University of Hawaii at Manoa, and Stanford University. For this group, the budgetary pressure brought by GPU shortages is particularly real:

"B3IQ's 'owner control' model is a viable path between renting and purchasing, which is exactly why we decided to partner with B3IQ. Research funding is fixed and allocated in advance, while cloud computing costs are fluctuating and might quietly eat up an entire budget line midway through a project. Turning compute power into a predictable, budgetable cost makes planning easier and makes it easier to account to project leaders or funding management departments." said Pavel Bushuyeu, AI researcher at the University of Hawaii. "Owning our own compute power also protects us from the impact of 'compute shortages'. When GPU supply is tight, centralized service providers allocate limits, and academic users often rank behind paying enterprise customers. With our own nodes, there is no need to queue and compete when tasks need to be run. And when the system is idle, using part of the idle compute earnings to offset the purchase cost can also amortize this investment."

For researchers and institutions that cannot hand over sensitive data to third-party model service providers such as Anthropic and OpenAI, this is a significant market. B3IQ pilot users like Pavel Bushuyeu from the University of Hawaii are running proprietary models on the platform for cancer research and robot training; this sensitive data simply cannot be transmitted externally. Additionally, AI service providers will directly block specific keywords and topics, potentially making entire research fields impossible to conduct: In Professor Yorke E. Rhodes III's Ethical Tech CoLab laboratory at the New York University Center for Global Affairs, master's students build frameworks and simulations based on war zone evacuation data and deduce national-level diplomatic negotiations. This kind of work would trigger content filtering by commercial models and can only be run on the team's own infrastructure.

B3 Labs believes this model can work now due to the rapid progress of open-source models: downloadable and self-deployable models are now powerful enough that institutions no longer need to call interfaces from centralized vendors like OpenAI; there is only one prerequisite: you must have your own machine to run them. This is exactly the link B3IQ fills.

"Institutions want better control over where AI runs, how data is processed, and how much cost is incurred for compute power." said B3 Labs CTO Sean Geng. "B3IQ brings these decisions into the same system: running private workloads on exclusive hardware, and then generating value from idle GPU compute power through a network that can be voluntarily joined."

Whether you want to purchase GPUs, rent out idle compute power, or directly rent compute power in the network, you can learn more at b3iq.org.

About B3 Labs

B3 Labs builds software and hardware for enterprise AI. Founded in 2024, the founding team comes from Coinbase, and the company has received over $21 million in investment from institutions such as Pantera Capital and Coinbase Ventures. B3 Labs operates two product lines: B3OS is an enterprise AI execution engine that allows AI agents to execute tasks stably and controllably within enterprise systems; B3IQ is GPU infrastructure fully owned by customers and deployed within the United States. For more information, please visit B3OS.org and B3IQ.org.