Seol mar théacs é seo: Efficiency and scalability of fully-resolved fluid-particle simulations on heterogeneous CPU-GPU architectures