Osegi
Startups·1 min read

Sparxity, formerly SpiNNcloud, raises €80 million for AI chips

Dresden-based Sparxity, formerly SpiNNcloud, has closed its Series A, lifting total funding to €80 million to build chips for sparse AI inference.

Server racks in a data center running specialised sparse-native AI computing hardware.
Image: EU-Startups

Dresden-based Sparxity, which has dropped its former name SpiNNcloud, has closed its Series A round, bringing cumulative funding to €80 million (about $90 million). The company builds sparse-native computing for AI and high-performance computing.

The round was led by ICM HPQC Fund, with Lockheed Martin Ventures, GlobalFoundries, GS Futures, Cloudberry Ventures, TGFS and the EIC Fund also taking part. Angel investors include Richard Socher and Nancy Xu, alongside continued backing from Atlantic and Silicon Catalyst.

What the money is for

The funding will go into specialised chips for fine-grained dynamic sparsity. Sparxity is targeting up to 44× faster AI inference in ultra-sparse models by computing only what matters. CEO and co-founder Hector Gonzalez says the aim is to move from dense, energy-intensive architectures to heterogeneous data centers with dedicated silicon for ultra-sparse workloads.

Sparse computing cuts unnecessary operations and data movement. On selected workloads, the company reports 10× faster sparse solvers, 18× lower energy use for event-based machine learning and 20× shorter runtimes for optimization problems. Its next-generation chip is being designed specifically for dynamic sparsity in AI inference.

Traction and context

Sparxity says it quadrupled revenue between 2024 and 2025, with momentum continuing in 2026. It points to three tapeouts, more than 70,000 chips produced, 12 racks deployed and over 30 customers worldwide. Recent milestones include a multi-million-dollar deployment at the Jülich Supercomputing Centre and a partnership with Deutsche Telekom to bring its technology into T-Systems data centers.

The round adds to a busy year for European AI hardware. OLIX raised €270.5 million for inference chips, Openchip €115 million, and EUCLYD announced a Series A above €200 million. The company founded in 2021 now positions itself as a European option for energy-efficient AI compute.

Apps worth a look

SUMO: the number puzzle

Tap the right tiles, add them up, hit the target. Easy to learn, hard to master. Free, for all ages.

Tags:sparxityspinncloudseries aai chipssparse computingdresden

More from this day

See all →

More in: Startups

See all →