TensorWave, a Las Vegas-based AI cloud provider specializing in AMD Instinct GPU infrastructure, announced a $350 million Series B funding round, at a $1.55 billion post money valuation.
TensorWave’s $350 million round was co-led by Magnetar and AMD Ventures, with continued participation from Maverick Silicon, Nexus Venture Partners, and Western Frontier. This follows rapid prior growth: a ~$43 million SAFE/seed round in October 2024 (the largest SAFE for a Nevada startup at the time) and a $100 million Series A in May 2025 (also led by Magnetar and AMD Ventures at a ~$400 million valuation). Total funding now approaches or exceeds ~$493 million.
What is TensorWave?
Founded in late 2023 by CEO Darrick Horton (Forbes 30 Under 30 AI honoree with data center, semiconductor, and prior FPGA cloud leadership experience), President/COO Piotr Tomasik, and Chief Growth Officer Jeff Tatarchuk, TensorWave is an AMD-exclusive “neocloud” focused on high performance, memory intensive AI workloads such as large language model training, fine tuning, high throughput inference, and generative AI.
It differentiates itself by avoiding NVIDIA dominance and vendor lock-in, emphasizing open ecosystems (AMD Instinct accelerators + ROCm), high bandwidth memory (HBM), liquid cooling, and expert support. Key offerings include bare-metal access, orchestration, storage, networking, and 24/7 monitoring with enterprise certifications (SOC II Type 2, ISO 27001, HIPAA).
The company operates one of North America’s largest AMD-based AI training clusters (8,192 MI325X GPUs, liquid cooled) and is scaling MI355X deployments. It has secured over 2 GW of long term data center capacity for multi region expansion, with headquarters in Las Vegas and hiring across engineering, operations, sales, and customer success.
Customers and traction include AI native firms like Moreh (frontier LLM inference on MI355X with significant optimizations), Fireworks AI, and Luma AI for large scale generative workloads and production inference.
The funding arrives amid explosive AI compute demand, where NVIDIA supply constraints create opportunities for alternatives. TensorWave bets on AMD’s Instinct lineup (MI300X, MI325X, MI355X, and upcoming models like MI455X) for strong memory capacity/bandwidth suited to massive models, combined with ROCm software and optimizations that enable competitive performance (e.g., CUDA compatibility layers showing substantial speedups in some cases).
Key advantages highlighted:
- Memory-optimized for frontier models: High HBM capacity and bandwidth for training/inference scalability.
- Open and flexible: Avoids proprietary lock-in; broad framework/library support.
- Production focus: Bare metal, expert engineering support, high utilization/reliability, and global scaling.
- Cost/performance: Competitive economics vs. constrained alternatives.
AMD Ventures’ involvement and quotes underscore strategic alignment: TensorWave helps expand AMD’s AI ecosystem footprint and provides an outlet for growing Instinct supply. Investors like Magnetar emphasize execution speed, reliability, and TensorWave’s role as a major AMD compute provider.
The ~4x valuation increase from the Series A reflects strong execution (cluster deployment, customer wins, capacity deals) in a hot market for specialized AI infrastructure, alongside broader “neocloud” momentum.
Recommended: Highland Electric Fleets Received $75 Million In Funding
How will TensorWave use the funds?
The $350 million will primarily fund:
- Global infrastructure expansion, including larger MI355X (and future) GPU clusters across new North American data center regions.
- Scaling operations, team growth (already tripled headcount post Series A), and Las Vegas headquarters.
- Meeting surging demand for memory intensive workloads while maintaining high reliability and support.
TensorWave positions itself for the “next phase of AI”, moving from experimentation to production at scale, by offering accessible, high capacity compute without long wait times or ecosystem lock-in.
Risks and Challenges
- Competition and tech execution: NVIDIA’s ecosystem dominance, software maturity gaps in ROCm vs. CUDA (despite compatibility progress), and rivals in AMD/multi vendor clouds.
- Capital intensity: Building/operating massive GPU clusters and data centers requires ongoing heavy investment; power, cooling, and supply chain execution are critical.
- Market dynamics: AI hype cycles, model efficiency improvements, or shifts in hyperscaler strategies could affect demand. Valuation multiples in private AI infra are elevated and sensitive to execution milestones.
- Operational scaling: Rapid growth from startup to major cluster operator demands strong delivery on reliability, utilization, and customer success.
This Series B marks a significant acceleration for TensorWave, validating its AMD-first thesis in a NVIDIA-heavy market. The combination of strategic AMD backing, proven cluster deployment, early customer traction with high profile AI firms, and substantial capacity pipeline positions it as a credible alternative provider for enterprises and AI labs seeking scalable, memory rich compute. Success hinges on continued execution in cluster scaling, software optimizations, utilization rates, and delivering differentiated price/performance/reliability. In the broader AI infrastructure boom, TensorWave exemplifies the rise of specialized players challenging incumbents through hardware alternatives and focused infrastructure plays.
Please email us your feedback and news tips at hello(at)techcompanynews.com
