Cerebras $2.5B: Physical AI Enters Manufacturing B2B

The timeline: from a solo bet to a $2.5 billion exit

In 2015, Lior Susan founded Eclipse Ventures with a contrarian premise. While American venture capital chased SaaS and digital marketplaces, Susan bet on the physical world. Robotics, semiconductors, industrial automation: sectors considered slow, capital-intensive, unglamorous.

Cerebras Systems was one of those investments. The company developed unusually large AI chips — the Wafer-Scale Engine — designed to accelerate the training and inference of large models. Furthermore, its positioning was explicitly an alternative to Nvidia: fewer commodity GPUs, more specialized architecture for intensive AI workloads.

In May 2026, TechCrunch has reported that Eclipse considers this exit just the beginning. The thesis on the physical world is not exhausted — on the contrary, it is proving more relevant than ever.

Architecture of the bet: why silicon matters more than software

Cerebras has built a real competitive advantage. Its chip integrates billions of transistors on a single wafer, eliminating communication latency between separate chips. Therefore, for applications requiring fast inference — in-line quality control, predictive maintenance, computer vision — this approach offers measurable benefits.

The distinction from traditional cloud AI is significant. Models hosted on remote infrastructure introduce latency, connectivity dependence, and data sovereignty issues. Conversely, an on-premise or edge architecture allows for local processing, which is crucial in industrial contexts where production data is sensitive.

Therefore, Cerebras's value is not just financial. It represents the maturation of an entire category: specialized AI hardware for physical applications. According to Gartner , by 2027 over 40% of enterprise AI workloads will be executed in edge or on-premise mode. The Cerebras deal anticipates this trajectory.

Winners and losers: who really profits from this scenario

The most obvious winner is Eclipse Ventures. The firm has shown that investing in the physical world is not a romantic niche, but a thesis with concrete returns. Furthermore, the credibility gained will attract capital towards other portfolio companies in AI hardware and robotics.

Cerebras itself is solidifying its position as a credible alternative to Nvidia. However, the road is still uphill: Nvidia controls over 70% of the AI GPU market according to industry estimates. Cerebras' specialization is an advantage in specific segments, not a generalized market conquest.

SMEs that postpone AI infrastructure decisions risk losing ground. In fact, while large industrial players — automotive, aerospace, pharmaceuticals — are already evaluating on-premise AI architectures, medium-sized companies risk finding themselves at a structural disadvantage. Consequently, the competitive gap could widen in the next 18-24 months.

The view from a Milanese agency: what we see in the Italian market

We at SHM Studio we work daily with Italian SMEs in the B2B and retail sectors. We observe a clear trend: the demand for AI solutions is growing, but often clashes with a still superficial understanding of the necessary infrastructure.

Many companies associate AI exclusively with cloud tools — ChatGPT, Copilot, SaaS platforms. However, for those operating in manufacturing, logistics, or supply chains, the real opportunity lies in AI applied to physical processes. This requires a focus on hardware, not just software.

The Cerebras-Eclipse deal is a market signal that even Italian SMEs should read. Therefore, those who deal with Digital marketing and digital transformation cannot ignore the infrastructural dimension of AI. The two things — digital communication and technological infrastructure — are increasingly interconnected.

Operational implications for Italian manufacturing SMEs

What does this scenario concretely mean for an Italian manufacturing company with 50-500 employees? First of all, it is useful to distinguish between three levels of AI maturity.

• Basic level: use of cloud AI tools for marketing activities, Copywriting , data analysis. Accessible today, with contained investments.
• Intermediate level: integration of AI models into specific business processes — predictive CRM, customer segmentation, campaign optimization Google Ads and Linkedin . Requires structured digital skills.
• Advanced level: on-premise AI for quality control, predictive maintenance, computer vision. Requires hardware investment and specific skills.

Furthermore, it is important to assess one's current position before jumping to advanced solutions. A company that has not yet optimized its SEO positioning or your digital presence is not ready for on-premise AI. Sequence matters.

For this reason, our approach in SHM Studio always starts with an assessment of overall digital maturity. Only then is it possible to identify where AI generates real value, and where it risks being a premature investment.

The ongoing construction site: where the game will be played in the next 24 months

Eclipse Ventures' thesis doesn't end with Cerebras. According to founder Susan, the portfolio includes dozens of companies operating at the intersection of AI and the physical world. Therefore, in the coming years, we will likely see other significant exits in robotics, energy tech, and industrial automation.

At a macro level, McKinsey estimates that generative and applied AI could add up to $4.4 trillion annually to the global economy. A significant portion of this value will come from industrial and physical applications, not just digital ones.

In Italy, the context is specific. The manufacturing fabric is mainly composed of SMEs with strong sectoral specialization. Therefore, the adoption of physical AI — even in less sophisticated forms than Cerebras chips — represents a real competitive opportunity. Especially for those who export and compete in international markets.

Similarly, anyone dealing with web development and digital presence must start thinking about how on-premise AI will change data flows and integrations between physical systems and digital platforms. The line between IT and OT (Operational Technology) is rapidly blurring.

Next moves: three priorities for those who want to get ahead

Based on the analysis, it is possible to identify three priority areas of action for Italian SMEs that want to prepare for this scenario.

First priority: digital maturity audit. Before evaluating any AI investment, it is necessary to have a clear vision of the current state. This includes SEO , web infrastructure, internal data quality, and analysis capabilities. Without this foundation, any AI investment risks being ineffective.

Second priority: internal training on applied AI. It's not about becoming experts in chips or hardware architectures. It's about understanding which business processes can benefit from AI, and what skills are needed to evaluate vendors. Furthermore, training reduces the risk of decisions based on hype rather than analysis.

Third priority: selection of technological partners. The B2B AI market is crowded with solutions. Choosing partners with vertical experience in your sector is crucial. Finally, it's useful to assess whether the proposed solutions are scalable – meaning they can grow with the company without requiring costly migrations.

To delve deeper into these topics or start an assessment of your digital position, you can contact the SHM Studio team or explore the contents of the our blog .