Alex Medick

TensorWave CEO Darrick Horton on open-source AI, AMD MI355X GPUs, and the future of enterprise AI infrastructure.

Can an Open-Source Approach Truly Democratize AI? Insights from TensorWave CEO Darrick Horton

Recently, SecureITWorld sat down with Darrick Horton, Founder & CEO of TensorWave, to discuss the future of AI infrastructure, the role of open ecosystems, and how AMD’s latest GPUs are reshaping what’s possible at scale.

From his early career working on classified projects at Lockheed Martin’s Skunk Works to leading one of the fastest-growing AI infrastructure companies today, Darrick brings a unique perspective on what it takes to push the boundaries of AI compute.

TensorWave has quickly mad

Recently, SecureITWorld sat down with Darrick Horton, Founder & CEO of TensorWave, to discuss the future of AI infrastructure, the role of open ecosystems, and how AMD’s latest GPUs are reshaping what’s possible at scale.

From his early career working on classified projects at Lockheed Martin’s Skunk Works to leading one of the fastest-growing AI infrastructure companies today, Darrick brings a unique perspective on what it takes to push the boundaries of AI compute.

TensorWave has quickly made headlines by deploying over 8,000 AMD Instinct MI325X GPUs and becoming one of the first cloud providers to roll out AMD’s new MI355X GPUs, designed to power large-scale AI training, fine-tuning, and inference workloads. In this interview, Darrick shares how open-source approaches, AMD’s hardware roadmap, and deep partnerships are enabling TensorWave to deliver enterprise-ready AI infrastructure without the tradeoffs of proprietary lock-in.


Key Takeaways from the Interview


1. Democratizing AI Through Open Ecosystems

Most enterprises face roadblocks when scaling AI: high costs, vendor lock-in, and overwhelming complexity. TensorWave was founded to change that. By building exclusively on AMD’s open ecosystem, the company delivers scalable, predictable, and cost-effective AI infrastructure at enterprise scale, without forcing customers into closed environments.


2. MI355X: A Game-Changer for AI Workloads

The newly launched AMD Instinct MI355X GPU offers 288GB of HBM3E memory and 8TB/s bandwidth, enabling fewer GPUs to do more work with faster runtimes. TensorWave sees the biggest performance gains for customers training large language models, powering enterprise copilots, and running high-throughput inference.


3. Scale and Specialization Matter

TensorWave’s 8,192 MI325X GPU cluster is the largest AMD-specific AI training infrastructure in North America. That scale gave TensorWave early access to the MI355X, ensuring their stack is tuned for AMD’s next-generation hardware. The result: enterprise-ready AI infrastructure that rivals—and often outperforms—proprietary alternatives.


4. GigaIO’s FabreX Eliminates Bottlenecks

To fully unlock GPU bandwidth, TensorWave integrates FabreX technology, which dynamically routes data across thousands of GPUs without congestion. This ensures customers can fully utilize the MI355X’s massive bandwidth, translating to faster training and more efficient inference.


5. Fueling Growth with $100M in Series A Funding

TensorWave’s $100M Series A, co-led by Magnetar and AMD Ventures, validates its AMD-first strategy and accelerates deployment of next-gen AI infrastructure. The funding will help expand clusters, optimize for future AMD architectures, and deliver enterprise-grade compute to a rapidly growing market projected to exceed $400B by 2027.


6. Strategic Partnerships Drive Real-World Results

Collaborations with MK1 (software optimization) and TECfusions (advanced cooling) enable TensorWave to maximize MI355X performance while improving efficiency and sustainability. A new cloud marketplace partnership will soon make AMD-optimized infrastructure broadly available.


7. Preparing for the Next Wave of AI

Looking ahead, Darrick highlights three big shifts:

 * Distributed AI across edge-to-cloud systems
 * Multimodal AI (text, vision, speech, more) becoming standard
 * Specialized hardware accelerators tuned for specific models

TensorWave’s AMD-exclusive focus and open approach uniquely position it to thrive in this evolving landscape.


Final Thoughts

Darrick Horton’s vision for TensorWave goes beyond simply building bigger GPU clusters. It’s about reshaping the economics, accessibility, and performance of AI infrastructure. By betting on open ecosystems, AMD’s roadmap, and strategic partnerships, TensorWave is proving that democratized access to cutting-edge compute isn’t just possible… it’s happening now.

👉 You can read the full interview with Darrick Horton on SecureITWorld.


About TensorWave

TensorWave is the AI cloud purpose-built for performance. Powered exclusively by Instinct™ Series GPUs, we deliver high-bandwidth, memory-optimized infrastructure that scales with your most demanding models—training or inference.

Ready to get started? Connect with a Sales Engineer.

TensorWave Welcomes the AMD Instinct™ MI355X

Can an Open-Source Approach Truly Democratize AI? Insights from TensorWave CEO Darrick Horton

Subscribe to our Blog

Stay ahead of the curve with the latest in AI, AMD accelerators, and all things TensorWave.

Product

Resources

Company

© 2025 TensorWave Inc. - All rights reserved.