Modeling Task Mapping for Data-intensive Applications in Heterogeneous Systems
This work provides a systematic modeling framework for task mapping in heterogeneous systems, addressing the challenge of optimizing data-intensive applications across diverse computing devices.
The paper introduces a new model for task mapping in heterogeneous systems (CPUs, GPUs, FPGAs) that accounts for inter-device communication and device-specific parallelizability. It demonstrates the model's utility through two novel mixed-integer linear programs, enabling systematic algorithm design.
We introduce a new model for the task mapping problem to aid in the systematic design of algorithms for heterogeneous systems including, but not limited to, CPUs, GPUs and FPGAs. A special focus is set on the communication between the devices, its influence on parallel execution, as well as on device-specific differences regarding parallelizability and streamability. We show how this model can be utilized in different system design phases and present two novel mixed-integer linear programs to demonstrate the usage of the model.