The EU-funded 1MICRON project targets ~1 µm spatial resolution using deep-silicon, edge-on X-ray sensors for phase-contrast imaging and intra-operative margin assessment. To achieve this, we require predictive simulations that connect X-ray interaction physics to charge transport and signal formation at micrometre pitch. We request a NAISS Small, CPU-only allocation on Dardel to run high-statistics simulations of charge-cloud propagation with explicit Coulomb repulsion in silicon. Objectives are to (i) quantify space-charge broadening and charge sharing at clinically relevant flux, (ii) define operating envelopes (bias, thickness, geometry) that preserve ~1 µm resolution, and (iii) provide design guidance for pixel layout and readout prior to fabrication and measurements. We will use Allpix Squared with Geant4 to generate X-ray interaction seeds and to propagate charge carriers with drift–diffusion while including Coulomb repulsion during transport. From simulated signals we will derive image-quality metrics (contrast-to-noise, MTF/NNPS proxies), 3D cloud descriptors (widths, tails, collection times), and task-based figures of merit relevant to soft-tissue contrast. Parameter studies will span spectrum, photon flux, sensor thickness, pixel pitch (~1 µm), and bias conditions representative of the 1MICRON design. This workload requires substantial CPU throughput: millions of events are needed to resolve trends across parameter grids and yield statistically robust recommendations. Dardel’s CPU partition supports thread-parallel Allpix/Geant4 runs and large job arrays, shortening iteration cycles with the hardware and reconstruction teams. The resulting models and datasets will de-risk prototypes, set safe operating points at clinical flux, and accelerate translation of the 1MICRON detector concept toward clinically useful imaging.