Nvidia (United Kingdom)

PIGNN3D: an accelerated physics-informed graph neural network for 3D thermal field simulation in data centers

Abstract Efficient thermal management is critical in data centers, where computational fluid dynamics (CFD) simulations provide high-fidelity airflow and temperature predictions but remain computationally demanding and time-intensive. While data-driven methods have emerged as pro...

Covariance-Guided Activation Selection with Adaptive Compensation for Efficient Neural Network Pruning

Structured neural network pruning physically deletes entire channels and intermediate fea- ture projections, delivering immediate memory and computational savings without requiring specialized sparse hardware accelerators. However, dominant channel selection heuristics— predomina...

Context-Conditioned Option Selection: A Black-Box Investigation of Jev and the Evolution of a Similarity-Based Decision Model

Specialised AI decision models select actions or answers from predefined alternatives given a context and a question. This paper examines whether the observed selection behaviour of Jev can be ap- proximated by context–question–option compatibility scoring followed by probability...

Interfacial Salinity–Transport Matching in 3D Solar Evaporators: A Framework for Brine‐Stable Solar Interfacial Evaporation

ABSTRACT Solar‐driven interfacial evaporation is commonly optimized through photothermal absorption and evaporation flux normalized by projected illuminated area. This metric is useful for quasi‐2D evaporators and dilute feeds, but it does not resolve the coupled constraints that...

Suburbanisation effects on light pollution in Warsaw’s natural areas assessed using drone imagery and machine learning

Suburbanisation is a major driver of artificial light at night (ALAN) in urban environments, yet its specific impacts on ecologically valuable areas remain insufficiently quantified. This study investigates the effects of suburbanisation on light pollution (LP) across eight ecolo...

Closed-form solutions for rigid transformations between heterogeneous groups of noisy geometric primitives

Abstract Computing the transformation between noisy collections of geometric primitives is a classic problem in three-dimensional computer vision and geometric computing. In this paper, we consider minimum sets of noisy objects and provide closed-form formulae to compute the Eucl...