Samsung, a global technology titan, has strategically invested in Euclyd, a Dutch artificial intelligence chipmaker, by participating in a significant $231 million Series A funding round. This move underscores the escalating demand for AI chip alternatives, challenging the current dominance of graphics processing units (GPUs) primarily associated with Nvidia.
Bernardo Kastrup, CEO of Euclyd, exclusively shared with CNBC that the company’s substantial 200 million euro funding round was co-led by Samsung, alongside prominent investors Somerset Capital Partners, the Scaleup Europe Fund managed by EQT, and Innovation Industries.
Founded in 2024, Euclyd is focused on developing a novel AI chip system. Its architecture diverges from traditional GPUs, specifically targeting the demanding requirements of AI inference, encompassing both the processor and memory subsystems. This strategic design aims to address the inherent limitations of current GPU-based solutions for the burgeoning AI ecosystem.
Nvidia’s meteoric rise to become the world’s most valuable company was largely fueled by the repurposing of its GPUs, originally designed for the gaming industry, into powerful tools for training and deploying AI models. However, this success has led to a near-monopolistic hold on the high-end chip market. Consequently, major hyperscalers and a growing number of startups are actively pursuing the development of their proprietary AI processors to optimize AI workloads and secure a more resilient supply chain.
This trend is further evidenced by announcements from industry leaders. OpenAI, for instance, revealed in August its first in-house AI chip, codenamed “Jalapeño,” claiming “industry-leading speed and efficiency.” Giants like Google, AWS, and Meta are also demonstrably investing heavily in their own custom AI silicon initiatives.
“AI is rapidly evolving into a foundational pillar for economic growth, scientific breakthroughs, and national competitiveness. However, its full potential will remain tethered unless we fundamentally re-architect the underlying infrastructure that powers it,” stated Kastrup.
While Euclyd’s systems are yet to be rigorously tested and proven in large-scale commercial deployments, the company asserts that its silicon solutions for foundational AI models will significantly reduce the energy consumption and operational costs associated with AI data centers – a sector currently experiencing a substantial surge in investment.
Euclyd has outlined a dual-pronged revenue strategy. Firstly, it plans to directly supply its hardware and integrated rack systems to enterprise clients requiring secure, self-hosted AI inference capabilities. Secondly, the company intends to license its intellectual property to other firms eager to develop their own custom chips leveraging Euclyd’s pre-existing technological advancements.
Kastrup emphasized the strategic value of Samsung’s involvement beyond mere capital infusion. “Samsung’s contribution extends far beyond financial support,” he explained. “As one of the world’s leading memory manufacturers, they bring invaluable engineering expertise, a deep understanding of complex systems, profound knowledge of the global supply chain, and an extensive network that will be critical to our success.”
Looking ahead, Euclyd has ambitious deployment targets. The company aims to commence the rollout of its physical chip systems in 2028, with a strategic objective to serve thousands of enterprise customers by 2030, as revealed by Kastrup. This timeline positions Euclyd to capitalize on the projected exponential growth of AI infrastructure demands in the coming years.
Original article, Author: Tobias. If you wish to reprint this article, please indicate the source:https://aicnbc.com/25716.html
