Exploiting parallelism in many-core architectures: a test case based on Lattice Boltzmann Models

In this paper we address the problem of identifying and exploiting techniques to increase the performance of large scale scientific codes on recent many-core processors. We consider as a test-bed a state-of-the-art Lattice Boltzmann (LB) model, that accurately reproduces the thermo-hydrodynamics of a 2D-fluid that obeys the equations of state of a perfect gas. The regular structure of Lattice Boltzmann algorithms makes it relatively easy to identify a large degree of available parallelism; the challenge is that of mapping this parallelism onto processors whose architecture is becoming more and more complex, both in terms of an increasing number of independent cores and – within each core – of vector instructions on longer and longer data words. We take as an example the Intel Sandybridge micro-architecture, that supports AVX instructions operating on 256-bit vectors; we address the problem of efficiently implementing the key computational kernels of LB codes – streaming and collision – on these family of processors, introducing several successive optimization steps and quantitatively assessing the impact of each of them on performance. Our final result is a production-ready code that we are already using for large scale simulations of the Rayleigh-Taylor instability. We analyze both raw performance and scaling figures, and finally compare with GPU-based implementations of the same class of simulation codes.