Boundless scales distributed GPU compute networks to support AI inference workloads.

by
Alyssa Davidson

Posted: July 14, 2026 10:26 AM Updated: July 14, 2026 10:26 AM

by Anastasia Oh

Edit and fact check date: July 14, 2026, 10:26 AM

We sometimes use automatic translation plugins to improve your local language experience. The automatic translation may not be accurate, so please read the original text for accurate information.

Boundless announced its expansion into the artificial intelligence sector with plans to deploy a distributed GPU network to support the growing demand for AI inference. The move expands the company’s existing infrastructure, originally developed to support zero-knowledge (ZK) proof workloads for the cryptocurrency industry.

“Four years ago, we set out to solve a hard computing problem,” said Shiv Shankar, CEO of Boundless. “Along the way, we built a network to coordinate something bigger: distributed GPU capacity. Now AI needs the same foundation at much larger scale. An open model gives teams more control over what they build, but production inference is still limited by cost, capacity, and reliability. Boundless is built to change that.”

The company’s ZK Proof Network is designed to coordinate distributed GPU resources, schedule compute-intensive tasks, and maintain efficient operations across multiple systems. Boundless said that same infrastructure is now being adapted to address the growing demand for AI inference as organizations scale their deployment of artificial intelligence applications.

According to the company, AI inference is emerging as a significant driver of computing demand, with infrastructure costs becoming an increasingly important consideration. Boundless, citing Gartner, estimates that 55% of spending on AI-optimized Infrastructure-as-a-Service will be allocated to inference in 2026, rising to more than 65% by 2029.

Endless repurposing of GPU networks for AI inference

Boundless says many AI inference workloads do not require premium data center GPUs and can instead run efficiently on inexpensive hardware, including consumer-grade graphics processors and GPUs originally purchased for compute-intensive applications such as cryptocurrency mining and ZK proof. The company said it aims to identify and optimize underutilized capacity by integrating hardware, software and inference service systems.

The company currently operates a distributed network of approximately 4,000 GPUs initially developed for ZK proofs. As part of its AI strategy, Boundless is optimizing its network to provide a managed inference infrastructure, including hardware optimization, workload adaptation, routing, batching, scheduling, monitoring, and operational support.

Boundless reports that initial benchmarking results show that inference costs can be reduced by up to 50% compared to hyperscale cloud providers, with the largest savings expected to be seen in asynchronous workloads. The company also plans to integrate the native ZKC token into the AI ​​Network, with operators expected to participate by staking their tokens and tying staking levels to earning potential.

Boundless said this expansion is intended to enable organizations to deploy AI inference workloads at lower compute costs. The transition to AI infrastructure was announced today, with full product launch scheduled for later this summer. The company added that it will continue to operate its zero-knowledge proof network along with its AI inference service.

As a dedicated journalist at MPost, Alisa specializes in the broad areas of cryptocurrency, AI, investing, and Web3. With a keen eye for new trends and technologies, she provides comprehensive coverage to inform and engage readers about the ever-evolving digital financial landscape.

• 
  
  

more articles