QumulusAI Inc (NASDAQ: QMLS) has secured a new GPU-as-a-Service agreement with DRW, a diversified global trading firm, to supply dedicated NVIDIA Blackwell B300 compute capacity. This deal marks the company’s second financial-markets customer, following an August 7 announcement with an agentic hedge fund, and adds to a book of business that now includes customer agreements valued at more than $246 million in aggregate since early June. The contract carries an initial one-year term with three one-year renewal options, allowing the arrangement to run up to four years in total.

The agreement enables DRW to expand the scale and speed of its compute resources using current-generation hardware served from QumulusAI’s U.S. data center footprint. DRW, which incubated Compute Exchange and Silicon Data to develop pricing benchmarks for the GPU compute market, will use the capacity to run complex research at greater scale. Rich Norman, Chief Information Officer of DRW, stated that the access allows the firm to iterate faster and shorten the distance between hypothesis and data insights.

Deal Structure and Strategic Context

Unlike the previous profit-sharing model with an agentic hedge fund, this agreement is structured as a standard GPU-as-a-Service contract. It provides DRW with fast, dedicated access to Blackwell compute without the variable revenue components tied to trading performance seen in earlier deals. Michael Maniscalco, CEO of QumulusAI, noted that quantitative trading firms are among the most demanding compute customers, requiring infrastructure that can keep pace with rapid market movements.

This development reinforces QumulusAI’s inference-first, demand-led deployment model, which aims to bring compute closer to customer demand rather than relying on traditional centralized hyperscale cloud models. The addition of DRW, a firm with deep expertise in commodities, derivatives, and emerging asset classes, signals growing institutional adoption of distributed AI infrastructure.

What the Numbers Show

The accumulation of over $246 million in announced customer agreements since early June demonstrates rapid traction for QumulusAI’s neocloud platform. By securing two distinct types of financial services clients—one with a variable profit-sharing model and another with a dedicated service contract—the company is diversifying its revenue streams within the high-performance computing sector. This dual approach allows QumulusAI to capture both fixed-value commitments and upside-linked opportunities, potentially stabilizing cash flows while maximizing the economic yield of its reserve assets.

QumulusAI (NASDAQ: QMLS), a neocloud infrastructure provider, will ring the opening bell at the Nasdaq MarketSite in New York's Times Square on Wednesday, July 29. The ceremony marks a public milestone for the company, which operates as a distributed AI cloud platform delivering accelerated access to high-performance GPU compute. The event underscores the firm's growing visibility in the AI computing sector as it scales its inference-first deployment model.

The live broadcast of the bell-ringing ceremony will begin at 9:15 a.m. EDT and can be viewed on nasdaq.com/marketsite/bell-ringing-ceremony. This public appearance serves as a recognition of the company's operational progress and market presence within the specialized cloud infrastructure space.

Executive Commentary

Michael Maniscalco, CEO of QumulusAI, highlighted the significance of the occasion for the organization and its stakeholders.

"Ringing the Nasdaq bell is a milestone moment in a company's life, and it is a privilege to mark ours," Maniscalco said. "We are grateful to the employees, board members, investors, partners and customers who helped us reach this point, and it means a great deal to celebrate it together on Nasdaq's stage."

Business Model Overview

QumulusAI differentiates itself through an inference-first, demand-led deployment model across a network of data center sites. This approach brings compute resources closer to customer demand, allowing AI teams and enterprises to scale production workloads with greater speed, flexibility, and control.

By combining rapid deployment with flexible private cloud infrastructure, the company aims to provide a more adaptable path for clients facing capacity constraints associated with traditional centralized and hyperscale cloud models.

Key Details

The announcement reinforces QumulusAI's position as a dedicated provider for the AI computing era, emphasizing its distributed architecture over traditional centralized alternatives.