A Domain Specific Approach to High Performance Heterogeneous Computing

TL;DR

This paper presents a domain-specific method for optimizing workload distribution in heterogeneous computing systems, improving performance and accuracy in financial modeling tasks through predictive modeling and advanced allocation techniques.

Contribution

It introduces a domain-specific approach that models application characteristics and uses these models for efficient workload allocation in heterogeneous systems.

Findings

Models predict workload latency and accuracy within 10% of actual performance.

Machine learning and MILP approaches significantly reduce latency, up to 270 times.

Demonstrated effectiveness in derivatives pricing with diverse hardware platforms.

Abstract

Users of heterogeneous computing systems face two problems: firstly, in understanding the trade-off relationships between the observable characteristics of their applications, such as latency and quality of the result, and secondly, how to exploit knowledge of these characteristics to allocate work to distributed computing platforms efficiently. A domain specific approach addresses both of these problems. By considering a subset of operations or functions, models of the observable characteristics or domain metrics may be formulated in advance, and populated at run-time for task instances. These metric models can then be used to express the allocation of work as a constrained integer program, which can be solved using heuristics, machine learning or Mixed Integer Linear Programming (MILP) frameworks. These claims are illustrated using the example domain of derivatives pricing in computational finance, with the domain metrics of workload latency or makespan and pricing accuracy. For a large, varied workload of 128 Black-Scholes and Heston model-based option pricing tasks, running upon a diverse array of 16 Multicore CPUs, GPUs and FPGAs platforms, predictions made by models of both the makespan and accuracy are generally within 10% of the run-time performance. When these models are used as inputs to machine learning and MILP-based workload allocation approaches, a latency improvement of up to 24 and 270 times over the heuristic approach is seen.