Financial Software on GPUs: between Haskell and Fortran,” in Funct

This paper presents a real-world pricing kernel for financial deriva-tives and evaluates the language and compiler tool chain that would allow expressive, hardware-neutral algorithm implementation and efficient execution on graphics-processing units (GPU). The lan-guage issues refer to preserving algorithmic invariants, e.g., inher-ent parallelism made explicit by map-reduce-scan functional com-binators. Efficient execution is achieved by manually applying a series of generally-applicable compiler transformations that allows the generated-OpenCL code to yield speedups as high as 70 × and 540 × on a commodity mobile and desktop GPU, respectively. Apart from the concrete speed-ups attained, our contributions are twofold: First, from a language perspective, we illustrate that even state-of-the-art auto-parallelization techniques are incapable of discovering all the requisite data parallelism when rendering th