TensorWave lands £100M to challenge AI compute bottlenecks with AMD-powered superclusters

AI infrastructure funding is ramping up as demand for powerful compute soars. TensorWave is among the latest to capitalise. The Las Vegas-based startup has secured £100M in Series A financing – a notable sum in a sector where some competitors have raised even larger rounds, but few have matched TensorWave’s capital efficiency. This funding will fuel the deployment of its AMD Instinct MI325X GPU-powered training clusters and support rapid operational scaling. Magnetar and AMD Ventures co-led the round, with participation fromMaverick Silicon, Nexus Venture Partners, and new investorProsperity7.
In an industry where access to AI compute often determines success or failure, TensorWave’s mission resonates with many founders and developers. Imagine your team has trained a promising large language model, only to find cloud GPU availability has dried up. TensorWave aims to change this scenario by offering AMD-powered infrastructure that is purpose-built for performance. This alternative effectively addresses capacity bottlenecks, especially as the AI infrastructure market is projected to reach $356 billion by 2032, with double-digit annual growth rates.
TensorWave’s leadership brings strong technical and entrepreneurial experience to the AI infrastructure. CEO Darrick Horton and President Piotr Tomasik hail from engineering and high-performance computing backgrounds. Tatarchuk and Tomasik were long-time pickleball doubles partners, while Horton and Tomasik were colleagues who worked on infrastructure optimisation projects. The founders bonded over a shared frustration – the limited and fragmented access to GPU resources for training frontier models.
Their shared vision – to build an AMD-optimised cloud platform specifically for high-bandwidth, memory-heavy AI workloads – sparked the founding of TensorWave. Their complementary expertise in hardware acceleration and cloud-native architecture helped accelerate the company’s early growth, attracting attention from top-tier investors and strategic partners.
At its core, TensorWave was founded to solve a problem that affects nearly every AI company today – the lack of reliable, high-performance GPU access. The founding team saw a future where enterprises couldn’t choose between performance and availability. With its new online cluster and fresh capital, TensorWave is one step closer to making that future a reality.
The newly raised AI infrastructure funding follows TensorWave’s deployment of 8,192 AMD Instinct MI325X GPUS, which form one of the largest dedicated training clusters in the market. The MI325X is AMD’s latest AI accelerator, featuring 256GB of HBM3E memory per GPU — a 33% increase over its predecessor — and delivering 6 TB/s peak theoretical memory bandwidth.
With 304 compute units and an 8192-bit memory bus, the MI325X is designed for the most demanding AI workloads, offering up to 40% faster inference for models like Mixtral 8x7B and 20% higher throughput for Meta Llama-3.1 70B compared to NVIDIA’s H100, according to AMD benchmarks. This deployment supports a growing roster of enterprise and AI-native customers. TensorWave expects this to push annual revenue run rate past $100M by year’s end [a 20x increase from last year], a testament to its capital efficiency compared to larger but less nimble competitors.
“This $100M funding propels TensorWave’s mission to democratize access to cutting-edge AI compute,” said CEO Darrick Horton. “Our 8,192 Instinct MI325X GPU cluster marks just the beginning as we establish ourselves as the emerging AMD-powered leader in the rapidly expanding AI infrastructure market.”
Co-founder and President Piotr Tomasik added, “The $100 million we’ve secured will transform how enterprises access AI computing resources. Through carefully cultivating strategic partnerships and investor relationships, we’ve positioned TensorWave to solve the critical infrastructure bottleneck facing AI adoption. Our Instinct MI325X cluster deployment isn’t just about adding capacity, it’s about creating an entirely new category of enterprise-ready AI infrastructure that delivers both the memory headroom and performance reliability that next-generation models demand.”
Jeff Tatarchuk, the company’s Chief Growth Officer, further emphasised their ambition: “Our focus is to expand the ecosystem and support developers with the tools, infrastructure, and performance they need to create, scale, and ship production-ready AI.”
Leading investor AMD Ventures brings not only capital but strategic alignment, as AMD continues to expand its AI accelerator portfolio.  “TensorWave is a key player in the growing AMD AI ecosystem,” said Mathew Hein, SVP Chief Strategy Officer & Corporate Development at AMD. “Their expanding portfolio of AI and enterprise customers coupled with their expertise in deploying AMD compute infrastructure is driving demand for access to their cutting-edge AI compute services. We’re excited to support their next phase of growth.”
Kenneth Safar, Managing Director at Maverick Silicon, echoed the sentiment: “We continue to be highly impressed by what the TensorWave team has built in just a short period of time. TensorWave is not just bringing more compute but rather an entirely new class of compute to a capacity-constrained market. We think this will be highly beneficial to the AI infrastructure ecosystem writ large, and we’re thrilled to continue our support of the company.”
Prosperity7 joins as a new investor, while Magnetar and Nexus Venture Partners returned with continued confidence in TensorWave’s growth trajectory. Nexus has a history of backing AI and infrastructure startups in global markets, while Magnetar brings experience in scaling capital-intensive technology platforms.
With over $100M in funding now secured, including earlier SAFE rounds, TensorWave plans to expand its team, invest in product development, and scale operations to serve more enterprises deploying complex AI systems. The company’s go-to-market strategy includes partnerships with leading cloud-native firms and research institutions.
As the global AI infrastructure market is projected to surpass $400B by 2027, competition is intensifying. Players like CoreWeave, Lambda Labs, and traditional cloud providers also target this segment. However, TensorWave differentiates itself through exclusive use of AMD Instinct GPUs and a focus on bandwidth- and memory-optimised architecture tailored for high-performance AI workloads. While CoreWeave recently secured $7.5 billion in debt financing and Lambda Labs raised $500 million to scale their NVIDIA-based offerings, TensorWave’s capital-efficient approach and technical focus allow it to punch above its weight.
“Our Instinct MI325X deployment isn’t just about capacity,” Tomasik noted. “It’s about enabling an entirely new generation of AI infrastructure to keep up with next-gen models. With AMD’s roadmap already pointing to even more powerful accelerators like the MI355X, our partnership positions us to remain at the forefront of this rapidly evolving sector.”