Sesame Summit 2026 – application open

Vertical Compute Raises €37M to Tackle AI’s Memory Bottleneck

The race to build the infrastructure underpinning artificial intelligence has moved well beyond software and into the physical layer of computing itself. As AI models grow exponentially in size and complexity, the semiconductor industry faces a fundamental constraint that threatens to slow progress: the memory wall. European deeptech investors are increasingly backing startups that address this hardware-level challenge, recognising that the next leap in AI performance depends as much on memory architecture as on processing power.

Belgian semiconductor startup Vertical Compute has raised an additional €37 million in seed funding, bringing its total seed round to €57 million. The company, an imec spin-off founded in 2024, is developing a novel 3D memory architecture that stacks vertical memory elements directly on top of compute logic within a single wafer manufacturing process. The fresh capital will fund the transition from chip validation to commercial chiplet deployment for next-generation AI platforms.

Quantonation Leads Extended Seed Round with Strong European Backing

The €37 million extension was led by Quantonation, the Paris-based deeptech venture firm specialising in quantum and advanced computing investments. New investors joining the round include Flanders Future Techfund (managed by Flemish investment company PMV), Wallonie Entreprendre, Sambrinvest, Noshaq, InvestBW, Drysdale Ventures, and Kima Ventures. Existing backers Eurazeo, XAnge, Vector Gestion, imec.xpand, and imec all reinvested, underscoring continued confidence in the company’s technical roadmap.

The breadth of the investor syndicate is notable. Vertical Compute has assembled a coalition that spans French, Belgian, and German government-backed funds alongside established European venture capital firms. This cross-border alignment reflects the strategic importance that European policymakers and investors attach to building sovereign semiconductor capabilities, particularly in AI-critical components where supply constraints continue to tighten.

The initial €20 million seed round, led by imec.xpand, had already positioned Vertical Compute as one of Europe’s most well-capitalised deeptech seed-stage companies. With the extension, the total €57 million seed round places the startup firmly among the continent’s largest pre-Series A raises in the semiconductor space.

Solving the Memory Wall: From Centimetres to Nanometres

The technical challenge that Vertical Compute addresses is well understood but stubbornly difficult to solve. Current AI chip architectures rely on a horizontal separation between processing units and memory, creating a bottleneck as data must travel centimetres across circuit boards. Static Random Access Memory (SRAM) offers speed but at prohibitive cost and limited density, while Dynamic Random-Access Memory (DRAM) provides greater capacity but consumes substantial energy and bandwidth. As AI workloads intensify, the scaling limitations of both memory technologies have become a binding constraint on system performance.

Vertical Compute’s approach — what the company describes as the “skyscraper model” — fundamentally reimagines this architecture. By integrating vertical data lanes directly on top of computation units within a single wafer process, the technology reduces data movement from centimetres to nanometres. The company claims its approach has the potential to outperform conventional DRAM in terms of density, cost, and energy efficiency, offering a step change rather than an incremental improvement.

In just over a year since spinning out of imec, Europe’s leading nanoelectronics research centre, Vertical Compute has grown to a team of 25 across Belgium and France, successfully taped out its first vertically integrated memory-on-logic test chip, and secured a manufacturing partnership with TSMC. The collaboration with the world’s largest semiconductor foundry validates that the vertical memory approach can be produced at industrial scale — a critical milestone for any chip startup seeking commercial traction.

European AI Chip Ambitions Gain Momentum

Vertical Compute’s raise arrives at a moment of heightened strategic urgency for the European semiconductor sector. Global AI chip demand continues to outstrip supply, with memory constraints emerging as a particular chokepoint. The European Chips Act, which aims to double Europe’s share of global semiconductor production to 20 per cent by 2030, has created a favourable funding environment for startups addressing critical gaps in the AI computing value chain.

The company was co-founded by CEO Sylvain Dubois, formerly of Google, and CTO Sebastien Couet, who brings deep expertise from his time at imec’s research labs. Their combination of commercial scaling experience and advanced semiconductor research positions Vertical Compute to bridge the gap between laboratory innovation and volume production — a transition that has historically proven challenging for European chip startups.

With the extended seed funding secured, Vertical Compute is now focused on moving beyond validation and into commercial chiplet deployment. If the technology delivers on its promise, it could offer AI system designers a European-made alternative to the memory solutions currently dominated by a handful of Asian manufacturers, whilst simultaneously addressing the energy consumption challenges that increasingly constrain data centre expansion.

Summary

CompanyVertical Compute
HeadquartersBelgium
Founded2024 (imec spin-off)
RoundSeed extension
Amount€37M (€57M total seed)
Lead InvestorQuantonation
Key InvestorsFlanders Future Techfund, Eurazeo, XAnge, imec.xpand, Kima Ventures
Use of FundsCommercial chiplet deployment for AI platforms
Key PartnershipTSMC manufacturing partnership

you might also like

FINTECH 1200x650 1
Fundraising 20 hours ago

London fintech Outpost raises $17.5M Series A led by Ribbit Capital to scale its AI-powered merchant-of-record platform, simplifying cross-border payments, tax, and compliance for global merchants.

AI fintech funding
Fundraising 2 days ago

The European fintech sector continues to attract early-stage capital, with AI-powered financial modelling emerging as a particularly active frontier for investor interest. As finance teams across high-growth organisations grapple with the limitations of static spreadsheets and fragmented planning tools, a new generation of startups is building intelligent infrastructure to replace legacy workflows. Stockholm-based Galdera Labs has now entered this space with a €1.5 million pre-seed round to develop an AI-native financial modelling platform designed for growth-stage finance teams. The funding will support platform development, reasoning infrastructure buildout, and an initial customer rollout targeting fast-growing companies with complex financial operations. Galdera’s platform combines a high-performance calculation engine with a semantic memory layer that links financial data directly to underlying business context, assumptions, and strategic decisions — enabling finance teams to query models in natural language and simulate complex scenarios in minutes rather than weeks. Klarna Veterans Back AI Financial Modelling Vision The pre-seed round was led by J12 Ventures, with participation from Antler and a roster of angel investors drawn from notable European technology companies including Klarna, DeepL, Stripe, and Plata. The investor composition reflects strong confidence in the founding team’s pedigree and the market opportunity for intelligent financial planning infrastructure. Galdera’s three co-founders — Evan Rumpza (CEO), Mattia Scolari (CFO), and Giovanni Casula (CTO) — met at Klarna during the fintech giant’s most intensive growth phase. Responsible for financial planning across 26 markets, the team experienced first-hand how manual processes and fragmented Excel models struggled to keep pace as business conditions shifted faster than traditional models could be rebuilt. To manage the complexity, they built an internal system at Klarna that replaced the static planning cycle with a continuously updated model — enabling what previously required large analyst teams to be handled by just three people, supporting the company through both capital raises and IPO preparations. The lessons learned from that experience became the foundation for Galdera Labs. “We’ve personally sat with 50 spreadsheets at two in the morning using tools that were supposed to solve the problem but didn’t. That is the infrastructure we are building with Galdera,” said Evan Rumpza, CEO and co-founder of Galdera Labs. Building AI Finance Tools for the Next Generation of CFOs The market for AI finance tools and financial modelling software is evolving rapidly as organisations demand more dynamic planning capabilities. Traditional spreadsheet-based approaches, while flexible, often create fragmented workflows where assumptions become outdated and institutional knowledge is lost between budget cycles. Galdera’s platform addresses this gap with a two-layer architecture: a powerful calculation engine capable of handling large data volumes, paired with a semantic memory layer that preserves the reasoning behind financial decisions over time. The platform is designed to function as an always-on financial forecast that automatically updates as business conditions change. Users configure scenarios once, and the model recalculates impacts across revenue, costs, margins, and other key metrics in real time. This approach positions Galdera within a growing wave of European fintech startups applying artificial intelligence not merely as an overlay on existing tools, but as a foundational redesign of how financial planning operates. With the launch, Galdera is opening its platform to its first customers: fast-growing companies and organisations with complex operations where the pace of decision-making has outgrown the tools finance teams traditionally rely on. Early adopters already include companies such as DeasyLabs, Unify, and Counsel. The pre-seed round positions Galdera Labs at an early but promising stage in a sector where demand for intelligent, context-aware financial infrastructure is accelerating across European markets. As AI continues to reshape enterprise workflows, the intersection of financial modelling and machine reasoning represents a significant opportunity for startups capable of delivering genuine operational value to scaling businesses. Summary

AevoLoop circular plastics recycling technology funding announcement with plastic waste processing
Fundraising 2 days ago

The sustainable consumer goods sector is witnessing growing investor appetite as environmentally conscious brands prove they can combine purpose with profitability. East London-based Allday Goods, the cult kitchen knife brand that transforms plastic waste into chef-quality blades, has raised £765,000 in a seed round led by FIGR Ventures to scale its operations from artisan favourite to mainstream kitchen staple. Founded in 2021 by ex-chef Hugo Worsley, Allday Goods manufactures kitchen knives with handles crafted entirely from recycled plastic waste — sourced from Maldon Salt buckets, milk bottle handles, discarded plant containers, and fishing nets washed up on British shores. The brand, which started in Worsley’s parents’ shed using a repurposed toastie maker, has already achieved profitability with minimal external investment. Products consistently sell out within minutes during online drops, and queues have formed at London pop-ups, reflecting a level of consumer demand that few sustainable brands can match at this stage. FIGR Ventures Leads Seed Round with Sustainability-Focused Backers The £765,000 round was led by FIGR Ventures, with participation from Anotherway Ventures, Machroes Holdings — the family office of Lord Mervyn Davies — and angel investor Tom Gozney, founder of the premium pizza oven brand Gozney. The investor mix signals confidence in Allday Goods’ ability to bridge the gap between sustainable manufacturing and scalable consumer product design. Allday Goods’ knives pair handles made from 100% recycled food-grade polypropylene with British and Japanese steel blades. The company collects, cleans, shreds, and remoulds plastic waste into distinctive, colourful handles that carry visible traces of their former lives — a design choice that has become central to the brand’s identity. Each knife effectively diverts plastic from landfill whilst delivering professional-grade performance. Worsley commented on the raise, noting that the team had built the brand slowly and intentionally, and that securing backing from investors they genuinely admire represents a significant milestone for the next chapter of growth. From Cult Following to Mainstream Market Opportunity Allday Goods has already demonstrated significant commercial traction without substantial marketing spend. The brand’s high-profile collaborations with Ottolenghi, Soho House, Maldon Salt, Kerrygold, and Paul Smith have positioned it at the intersection of culinary craftsmanship and design culture. Features in The World of Interiors and Esquire have further cemented its reputation among discerning consumers who value both aesthetics and environmental responsibility. The fresh capital will be deployed to scale production capacity, expand the product range, and accelerate the transition from limited-edition drops to consistent retail availability. The challenge for Allday Goods will be maintaining the artisan quality and brand mystique that fuelled its cult status whilst meeting the demands of a broader consumer base — a tension that many direct-to-consumer brands have struggled to navigate. The broader sustainable kitchenware market continues to attract both consumer interest and investor capital across Europe. As regulatory pressure on single-use plastics intensifies and consumers increasingly seek products that align with their environmental values, brands like Allday Goods that demonstrate genuine circularity in their manufacturing processes are well-positioned to capture meaningful market share. Summary Company: Allday GoodsHeadquarters: East London, United KingdomFounded: 2021Founder: Hugo WorsleyRound: SeedAmount: £765,000Lead Investor: FIGR VenturesOther Investors: Anotherway Ventures, Machroes Holdings, Tom GozneyUse of Funds: Scale production, expand product range, transition to mainstream retail availability

Subscribe to
our Newsletter!

Stay at the forefront with our curated guide to the best upcoming Tech events.