On June 10, 2026, a Las Vegas-based cloud computing startup announced a $350 million Series B funding round, the largest in Nevada history, at a post-money valuation of $1.55 billion. The raise itself was headline-worthy. What made TensorWave’s announcement genuinely remarkable was the principle at its core: the company had built an entire AI infrastructure empire without using a single Nvidia chip.
In an industry where Nvidia controls an estimated 70 to 80 percent of the AI accelerator market and where the waiting list for its H100 and H200 GPUs stretched months during the AI build-out of 2024 and 2025, TensorWave’s 28-year-old CEO Darrick Horton made a counterintuitive bet. He believed Nvidia controlled too much of the AI infrastructure market, considered the concentration harmful to competition, and designed TensorWave from the ground up around AMD hardware instead. Two years later, investors co-led by Magnetar and AMD Ventures handed him $350 million to prove the thesis at scale.
For Las Vegas, the story is bigger than one company. TensorWave’s ascent signals something that would have seemed implausible five years ago: the city that built its economy on hospitality and gambling is producing venture-scale technology businesses that compete at a global level.
From Seed to Unicorn in Under Two Years
TensorWave’s fundraising trajectory is unusual even by the frenzied standards of AI investment. The company raised a $43 million seed round in October 2024. It followed that with a $100 million Series A in May 2025, led by the same investors who returned for the Series B. Less than 13 months after the Series A, TensorWave closed $350 million more at a valuation that nearly quadrupled from where it stood at the Series A.
The new capital will support the expansion of TensorWave’s global AI infrastructure footprint, including the deployment of next-generation AMD Instinct MI355X GPU clusters designed for memory-intensive workloads such as large language model training, high-throughput inference, and generative AI applications.
The company is not theoretical. TensorWave now operates one of the largest AMD-based AI training clusters in North America, with 8,192 AMD Instinct MI325X GPUs online, and is preparing larger MI355X deployments across several new data center regions in North America. Those are operational GPUs running production AI workloads for paying customers including Fireworks AI and Luma AI, not research prototypes or vaporware projections.
TensorWave has also secured more than 2 gigawatts of long-term data center capacity to support growing adoption from enterprise, research, and AI-native customers. Two gigawatts of committed data center capacity, at a time when power availability is the binding constraint on AI infrastructure expansion globally, represents a significant strategic asset.
The pace of growth compelled Magnetar’s confidence. “The race to build AI infrastructure has created urgent demand for providers who can deliver at speed without sacrificing reliability,” said Ross Laser, Co-Founder and President of Magnetar. “TensorWave’s partnership with AMD and its disciplined execution make it exactly that, and we believe it is positioned to become one of the most important compute providers for AMD-based AI workloads.”
The Nvidia Problem That Created an Opportunity
To understand why TensorWave exists, you need to understand the structural challenge that anyone building AI infrastructure faced in 2023 and 2024. Nvidia’s GPU dominance created a bottleneck that no amount of money could instantly resolve. Companies that needed compute couldn’t get it quickly because Nvidia’s manufacturing pipeline couldn’t keep pace with demand. And even when hardware was available, customers were locked into Nvidia’s CUDA software ecosystem in ways that made switching painful.
Horton saw this concentration as both a business risk and a market opportunity. If AI builders were forced to rely on a single supplier with constrained supply and locked-in software, the provider willing to offer a credible alternative with reliable availability stood to capture enormous demand.
AMD’s Instinct series GPUs offered the hardware foundation. AMD had invested heavily in ROCm, its open-source software platform designed to compete with CUDA, and was making genuine progress on software compatibility that earlier generations of AMD accelerators had lacked. TensorWave bet that AMD’s hardware-software combination had crossed a threshold where it could support production AI workloads at scale.
AMD’s direct investment through its venture arm is notable. It signals that AMD views TensorWave as more than just a customer, but as a strategic channel for proving that AMD silicon can compete at scale in production AI workloads. When your primary chip supplier puts money in your company, the alignment of incentives runs deep. AMD benefits from TensorWave succeeding because every TensorWave customer is a proof point that AMD hardware can handle demanding production workloads.
In February 2026, TensorWave established a partnership with Credo, a networking technology company, to improve network reliability for large-scale GPU clusters. Networking is often the overlooked bottleneck in AI training. Large model training requires moving massive volumes of data between GPUs at extremely high bandwidth. Addressing networking reliability proactively signals operational sophistication beyond just acquiring and deploying hardware.
The Ambition Behind the Numbers
Horton’s stated targets would seem audacious from anyone else. He plans to have 100,000 GPUs deployed by next year, with hopes to eventually become the first company to deploy 1 million GPUs, a feat no company has ever achieved.
One million GPUs would represent a fleet larger than any publicly disclosed AI infrastructure deployment in history. For reference, Microsoft’s announced investment in OpenAI infrastructure and Meta’s disclosed AI hardware buildout are measured in tens of thousands to low hundreds of thousands of GPUs. A million-GPU deployment would be genuinely unprecedented.
The long-term vision is to provide a sovereign AI cloud that can support the training of models with trillions of parameters. While initial deployments in 2025 and early 2026 are focused on the 10-to-20 megawatt range, every milestone TensorWave has set so far, it has hit ahead of schedule.
The workforce implications are substantial for Las Vegas. TensorWave expects to double the number of employees over the next 12 months, from about 160 to between 300 and 400. Roles span engineering, infrastructure, operations, sales, and customer success, the full organizational stack required to operate a global GPU cloud at scale.
Those 150 to 240 new jobs in Las Vegas are not hospitality positions or service roles. They are engineering, infrastructure, and technical sales positions that typically command salaries well above the local median. For a metro area still working to diversify its economic base beyond tourism and gaming, TensorWave’s hiring plans represent meaningful structural progress.
The UNLV Connection and Local Roots
TensorWave’s co-founder and president, Piotr Tomasik, is a UNLV graduate. The company maintains its global headquarters in Las Vegas at Town Square, the outdoor shopping and entertainment center in the southern part of the valley. These local roots matter beyond symbolism.
A Las Vegas company founded in part by a UNLV graduate that grows to a $1.55 billion valuation creates a proof point that functions as recruiting and ecosystem-building material. Other UNLV graduates and Las Vegas residents watching TensorWave’s trajectory understand that building a significant technology company from Las Vegas is possible in ways that prior generations of local talent might have doubted.
Las Vegas ranked second in the world in startup community activity in recent rankings, a metric that would have seemed implausible even five years ago. TensorWave did not create that trend by itself, but its profile accelerates it. When a company from your city raises Nevada’s largest-ever Series B, it changes the conversation about what is buildable locally.
The decision to keep headquarters in Las Vegas rather than relocating to San Francisco or Seattle after raising at scale also signals something. Most well-funded tech startups face pressure to move to established tech hubs where investor density, talent pipelines, and corporate ecosystems are thicker. TensorWave’s choice to stay, and to expand hiring in Las Vegas specifically, reflects a deliberate commitment to building a technology company rooted in Nevada.
The Broader AI Infrastructure Race
TensorWave operates in a market that is growing rapidly but also attracting enormous competition. Hyperscalers including Amazon Web Services, Google Cloud, and Microsoft Azure are all expanding AI-focused compute offerings. Specialized GPU cloud providers including CoreWeave, Lambda Labs, and others compete in the same space as TensorWave.
What differentiates TensorWave is the AMD-exclusive positioning, which is genuinely unusual in a field where most competitors either use Nvidia hardware or offer a mix. This differentiation creates specific advantages and limitations. Customers who need AMD-based compute for software compatibility reasons, cost reasons, or supply availability reasons have a compelling reason to choose TensorWave over alternatives. Customers locked into Nvidia’s CUDA ecosystem face switching costs that may outweigh availability or pricing benefits.
The open-ecosystem argument may prove more important than it initially appears. Regulatory attention to Nvidia’s market position has grown alongside its market share. If antitrust scrutiny eventually creates pressure on Nvidia’s integrated hardware-software stack, providers offering credible alternatives benefit disproportionately. TensorWave’s AMD positioning would be strategically valuable in that environment in ways that are difficult to fully price today.
The power infrastructure angle also warrants attention. TensorWave has secured more than 2 gigawatts of long-term data center capacity. In a market where power availability has become the primary constraint on AI compute expansion, long-term power commitments represent durable competitive advantages that are difficult for new entrants to replicate quickly. Nevada’s energy infrastructure and relatively favorable regulatory environment for data center development make Las Vegas a workable location for this kind of power-intensive build-out.
What This Means for Las Vegas
TensorWave’s raise breaks Nevada’s Series B record and does so in a sector entirely disconnected from the casino economy that has historically dominated the state’s business identity. That symbolic break matters as much as the capital itself.
The funding validates a thesis that a small but growing community of Las Vegas technologists has been building toward: that the city’s tax advantages, affordable operations costs relative to coastal markets, growing talent base, and increasingly connected infrastructure make it a viable headquarters for technology companies that compete globally.
NV Energy’s 2026 Integrated Resource Plan, recently filed with state regulators, will govern how Nevada meets power demand over the next 20 years. TensorWave’s 2 gigawatts of committed data center capacity, alongside other AI and technology investment in the state, will be a significant input into that planning process. The feedback loop between technology sector growth and infrastructure planning creates compounding effects that reinforce Las Vegas’s position as an emerging technology hub.
The Golden Knights naming Richtech Robotics its “Rookie of the Year” partner the same week TensorWave announced its funding round was coincidental, but the juxtaposition captures something real. AI-driven robotics companies and AI cloud infrastructure companies are both choosing Las Vegas as their home. The technology ecosystem is broadening from individual outliers into something that looks more like a coherent cluster.
Key Takeaways
- TensorWave raised $350 million in Series B funding on June 10, 2026, the largest Series B in Nevada history
- The post-money valuation of $1.55 billion nearly quadrupled TensorWave’s value from its Series A just over a year earlier
- The round was co-led by Magnetar and AMD Ventures, with continued participation from Maverick Silicon, Nexus Venture Partners, and Western Frontier
- TensorWave operates exclusively on AMD hardware, deliberately avoiding Nvidia chips entirely
- The company operates one of the largest AMD-based AI training clusters in North America, with 8,192 AMD Instinct MI325X GPUs currently online
- TensorWave has secured more than 2 gigawatts of long-term data center capacity globally
- The company plans to grow from 160 to 300 to 400 employees over the next 12 months, all roles based at Las Vegas headquarters
- CEO Darrick Horton has set a target of 100,000 GPUs deployed within a year and eventually 1 million GPUs, a milestone no company has achieved
Important Insights
AMD’s decision to co-lead the Series B through its venture arm transforms the investor relationship into a strategic partnership. AMD is funding a company that functions as a live proof-of-concept for its hardware’s ability to compete with Nvidia at scale in production environments. TensorWave’s customer success becomes AMD’s marketing. This alignment creates incentives for AMD to support TensorWave’s growth in ways that go beyond financial returns on the investment.
The power infrastructure commitment of 2 gigawatts deserves more attention than it typically receives in coverage focused on valuation and GPU counts. Power availability is the real bottleneck in AI infrastructure expansion, not capital or hardware alone. Companies that lock in long-term power agreements in 2025 and 2026 are positioning themselves for competitive advantage in 2028 and 2030 when power constraints will be even more acute. TensorWave’s foresight here is a genuine strategic asset.
The founder’s anti-Nvidia conviction is not just philosophical differentiation. It is a structural bet that market concentration creates fragility and that fragility creates opportunity. Every major AI infrastructure buyer that worries about single-vendor dependency is a potential TensorWave customer. That concern is widespread enough in enterprise technology procurement to represent a large addressable market even without winning customers who have no Nvidia concerns at all.
Horton’s 100,000 GPU target for next year and eventual 1 million GPU aspiration are audacious, but TensorWave has consistently hit its stated milestones ahead of schedule. The discipline to execute on aggressive commitments rather than simply announce them is the rarer and more valuable capability. Investors who have watched the company deliver across three funding rounds in under two years have earned reason to take the targets seriously.
The UNLV connection and deliberate choice to remain Las Vegas-headquartered through multiple funding rounds represents a template for how local talent and local companies can build at global scale without abandoning the communities where they started. That template is more valuable for Las Vegas’s long-term technology ecosystem development than any single funding round.
For more information on TensorWave’s technology and services, visit TensorWave. For Nevada startup ecosystem resources, visit StartUpNV and the Nevada Governor’s Office of Economic Development.
