Summary

• Openchip has secured €115 million from SETT to develop chips for AI and high performance computing.
• The Barcelona company is working on RISC-V based systems and full stack AI and HPC compute.
• The investment links Spain’s technology policy to Europe’s wider push for semiconductor autonomy.

Openchip has secured €115 million from Spain’s Society for Technological Transformation, adding another state backed piece to Europe’s attempt to build more of the silicon stack needed for AI and high performance computing.

The Barcelona based company designs high performance, energy efficient chips and compute systems for AI and HPC workloads, including work around RISC-V based accelerators and systems on chip. The financing from SETT is intended to support chip development as European governments try to reduce dependence on foreign compute technologies without pretending that semiconductor autonomy can be delivered by one flagship project.

Openchip describes itself as a full stack system on chip and software company, combining in house chip design with software and systems integration. That full stack claim is important. In AI infrastructure, value is not captured only by the chip. Performance depends on hardware architecture, software tooling, memory, networking, power efficiency, and the ability to fit compute systems into real deployment environments.

The investment comes as Europe’s semiconductor ambitions are being tested by US China technology competition, export controls, and the high cost of advanced chip production. Europe has world class assets in parts of the supply chain, but its AI infrastructure still depends heavily on overseas chip designers, hyperscale cloud providers, and global manufacturing capacity.

Openchip’s RISC-V work gives the story a sharper policy angle. RISC-V is an open instruction set architecture that can allow companies and research institutions to design processors without relying on proprietary architectures controlled by a small number of global technology firms. Open hardware does not remove the difficulty of building competitive chips, but it gives European companies a route to experiment with more customised, energy aware designs.

AI compute is becoming both an industrial and an energy question. Model training and inference require specialised hardware, and the cost of running AI systems is increasingly shaped by power consumption as much as raw performance. Energy efficient chips for AI and HPC can reduce operating costs, support dense computing environments, and make European compute projects less exposed to imported hardware bottlenecks.

Spain’s role is also worth noting. European semiconductor policy is often discussed through Germany, France, the Netherlands, and Ireland, but the geography of chip strategy is widening. Public capital from SETT gives Spain a more visible position in the European digital sovereignty debate, particularly where AI, HPC, and open architectures overlap.

Openchip alone will not transform Europe’s compute landscape. The more realistic question is whether public investment can help create a deeper supplier base for specialised compute, giving European cloud providers, research centres, industrial companies, and public sector buyers more options over time.

The execution challenge is severe. Chip companies require long development cycles, expensive validation, manufacturing partners, customer commitments, and software ecosystems that developers are willing to use. Europe has a strong research base, but turning semiconductor research into commercially deployed infrastructure has repeatedly proved difficult.

Openchip’s full stack approach is relevant because AI hardware companies that focus only on silicon risk being trapped between dominant GPU ecosystems and hyperscale cloud purchasing power. Companies that combine chips, systems, and software have a better chance of addressing specific workloads where efficiency, sovereignty, and integration matter more than generic benchmark leadership.

The investment also sits inside a wider rethink of industrial policy. Europe’s digital economy increasingly depends on infrastructure it does not fully control: cloud platforms, chips, AI models, software stacks, and network equipment. Public money cannot eliminate that dependence quickly, but it can support strategically useful alternatives if governments are disciplined about demand, procurement, and scale.

Openchip now has to turn a sovereignty argument into deployable technology. That means proving performance, securing manufacturing and packaging routes, building software support, and finding customers with enough workload specificity to value a European alternative. The €115 million investment gives the company room to move; it does not remove the brutal economics of the semiconductor market.