GPU PRIME LTD

Their core work

GPU Prime is a Cambridge-based computational SME that applies high-performance, GPU-accelerated simulation and modelling methods to complex scientific and engineering problems. Their documented work spans molecular dynamics simulations of biological systems — specifically drug resistance mechanisms in tuberculosis — and aerodynamic and propulsion modelling for next-generation aircraft designs. The common thread across these seemingly unrelated domains is a capacity to run computationally intensive simulations that would be impractical on standard hardware. They join specialist research consortia as a focused computational partner, contributing modelling capability rather than domain-specific laboratory or manufacturing infrastructure.

What they specialise in

How they've shifted over time

GPU Prime's first H2020 engagement in 2019 was firmly in computational biology — molecular dynamics, metabolic pathway modelling, and fitness landscape analysis applied to tuberculosis drug resistance. By 2020, with no overlap in topic, they had joined an aerospace RIA project modelling disruptive electric propulsion architectures for aircraft. This is not a pivot in domain expertise but rather evidence of a domain-agnostic simulation capability being deployed wherever computationally intensive problems need solving — the methodology travels, the application area changes.

GPU Prime appears to be moving from niche computational biology toward larger-budget engineering simulation, suggesting an ambition to position as a cross-domain HPC modelling partner rather than a narrow life-sciences specialist.

How they like to work

GPU Prime has participated exclusively as a consortium partner — never as coordinator — across both H2020 projects. Despite just two projects, they were placed in well-populated, international consortia totalling 29 unique partners from 12 countries, which suggests they were specifically recruited for a computational contribution others in the consortium could not provide. Working with them likely means contracting a focused technical team that delivers simulation outputs as one specialist node in a larger research effort, with no expectation of project management or administrative leadership.

With only two projects, GPU Prime has established connections with 29 distinct consortium partners across 12 countries — an unusually wide network for such a small portfolio, indicating placement in well-connected international consortia. Their Cambridge location gives them proximity to one of Europe's densest clusters of deep-tech and computational research talent.

What sets them apart

GPU Prime is a rare SME that has contributed computational modelling to two entirely different scientific domains — infectious disease biology and aerospace engineering — within a three-year window. This cross-domain versatility, almost certainly grounded in GPU-accelerated simulation capacity, makes them a potentially valuable consortium partner when a project needs serious computational firepower without the overhead of a large research institute. For consortium builders, they represent an agile specialist who can integrate into diverse scientific teams wherever simulation bottlenecks exist.

Highlights from their portfolio

• ENODISE
  Their largest project by far (EUR 350,000), addressing boundary layer ingestion and distributed electric propulsion — technologies central to aviation's decarbonisation roadmap and attracting significant industry interest.
• AMR-TB
  A highly specific MSCA-RISE project applying GPU-intensive molecular dynamics to map how tuberculosis bacteria evolve resistance to first-line drugs — a high-impact public health problem tackled through a purely computational lens.