Auddia Inc. (NASDAQ: AUUD) has highlighted newly available market data to frame the scale of the AI infrastructure opportunity for its subsidiary LT350, a distributed AI data center company. The data comes from a recent SEC filing by SharonAI Holdings Inc. (NASDAQ: SHAZ), which disclosed a $1.25 billion, 60-month contract to deploy and operate 8,200 NVIDIA B300 GPUs for a global cloud customer. According to the filing, the contract equates to approximately $30,488 per GPU per year.
Auddia is using this benchmark to help investors contextualize the potential of LT350's distributed AI infrastructure model relative to a previously disclosed internal discounted cash flow analysis of $250 million for McCarthy Finney, the new holding company that will be formed after the merger of Auddia and Thramann Holdings, LLC. Fifty percent of that valuation was attributed to LT350. The new market pricing data suggests that LT350's market opportunity may be larger than initially considered.
LT350 is one of three new businesses—alongside Influence Healthcare and Voyex—that will be combined with Auddia to form McCarthy Finney, which will trade under the symbol NASDAQ: MCFN. The merger is expected to close pending shareholder approval and other conditions.
Jeff Thramann, CEO of Auddia and founder of LT350, emphasized the significance of the benchmark. "We are not providing forecasts or forward revenue projections outside of what is disclosed in our recent S-4 financing," he said. "But I believe publicly disclosed market pricing helps investors understand why distributed AI infrastructure is emerging as one of the most compelling opportunities in the sector."
LT350's business model is similar to SharonAI's, but with a key advantage: its proprietary datacenters are placed directly next to customers in defense, healthcare, financial services, and government sectors to optimize data sovereignty and latency. "Whereas SHAZ and others tout data sovereignty by keeping customer data within a country, LT350's key competitive advantage keeps customer data behind their firewall," Thramann added.
Auddia previously disclosed that LT350's REIT partner controls 4,000,000 square feet of suitable parking-lot airspace. LT350's patented canopy architecture is designed around 2,000-square-foot modules, each capable of supporting 480 GPUs using a 2:1 GPU-to-battery cartridge ratio. If fully deployed across the REIT footprint, LT350's architecture could support approximately 2,000 canopies, 480 GPUs per canopy, and 960,000 GPUs total capacity. Using the publicly disclosed benchmark of $30,488 per GPU per year, this footprint would represent $29 billion in annualized market-equivalent pricing. However, Auddia emphasizes that actual pricing, utilization, and deployment levels will vary and that the company applies significant discounts in its internal analyses.
Beyond the REIT partnership, LT350's canopy architecture is designed for broad applicability across hospitals, universities, retail centers, convenience stores, industrial facilities, municipal properties, stadiums, and multifamily housing. Key features include deployment in existing parking lot airspace, closed-loop liquid cooling with zero water consumption, battery-buffered operation for grid support, circuit-level grid deployment without transmission upgrades, modular scaling, and distributed mesh connectivity for edge compute with hyperscaler interoperability.
"The REIT footprint is just one example of how LT350 can scale," Thramann said. "Our IP portfolio gives us a proprietary position in a category that is only beginning to emerge." LT350 holds 13 issued, 1 allowed, and 2 pending patents on its solar parking lot canopy infrastructure platform. For more information, visit www.LT350.com. The company's whitepaper is available here.
