German electronics SME specialising in FPGA design and HPC hardware acceleration, with applied experience in deep learning on heterogeneous compute platforms.
PRO DESIGN Electronic GmbH is a German electronics engineering SME specialising in high-performance embedded computing and FPGA-based hardware design. In EU research projects they bring the practical hardware engineering skills needed to turn architectural research into working prototype systems — bridging the gap between theoretical HPC design and real silicon implementation. Their work spans low-level power-performance optimisation on manycore chips all the way up to accelerating deep learning inference on heterogeneous hardware platforms. For a consortium, they represent the "build it and make it run" function that turns academic ideas into demonstrable results.
MANGO (2015–2019) placed PRO DESIGN inside a consortium explicitly exploring manycore architectures for next-generation HPC, with a focus on partitionability, reconfigurability, and real-time determinism.
MANGO keywords (power-performance-predictability, capacity computing) indicate hands-on work optimising energy efficiency and deterministic throughput on complex compute platforms.
DeepHealth (2019–2022) brought PRO DESIGN into the AI domain, contributing heterogeneous architecture expertise to accelerate deep learning pipelines for biomedical applications.
Both projects involve heterogeneous or manycore compute paradigms, suggesting consistent capability in integrating CPUs, FPGAs, and accelerators into unified systems.
Between 2015 and 2019, PRO DESIGN's H2020 work centred on foundational HPC infrastructure — manycore chip architectures, real-time guarantees, and the hard problem of making high-capacity compute systems both reconfigurable and power-efficient. From 2019 onward the focus shifted toward applying those hardware skills to AI workloads: deep learning inference and the heterogeneous architectures (GPU, FPGA, CPU mixtures) needed to run it efficiently at scale. The trajectory is a logical upgrade path — from building fast, flexible compute platforms to putting those platforms to work for machine learning — rather than a pivot into a completely different domain.
PRO DESIGN is moving from research-grade HPC infrastructure toward AI-on-hardware applications, making them a relevant partner for any consortium that needs to run machine learning workloads on resource-constrained or real-time embedded platforms.
PRO DESIGN has participated in both of their H2020 projects as a consortium member rather than a coordinator, indicating they operate as a specialist technical contributor rather than a project leader. With 29 distinct partners across just two projects, they have been embedded in large, diverse consortia — suggesting they are comfortable working within complex multi-partner structures and contributing a defined technical workpackage. This profile fits an organisation that prefers to deliver a concrete engineering component rather than manage overall project governance.
PRO DESIGN has built a surprisingly broad network for a two-project SME, accumulating 29 unique partners across 10 countries through MANGO and DeepHealth alone. Their collaborations span research institutes, universities, and industrial partners across Europe, reflecting the large international consortia typical of FET and ICT-level projects.
PRO DESIGN occupies a rare niche as a hardware-focused SME that can contribute FPGA and embedded design expertise to research-grade HPC and AI consortia — a role usually filled by large semiconductor firms or university labs. Their participation in both a FET research project (MANGO) and an Innovation Action (DeepHealth) shows they can operate at both the exploratory research end and the application-delivery end of the funding spectrum. For a consortium coordinator, they offer credible industry participation with genuine technical depth rather than token SME presence.
