Adaptive Parallelization For Multi-scale Simulation

A method for optimizing turnaround time for a multi-scale simulation of material cases for use in semiconductor technology, comprising the steps of:

• a computer system receiving a case/task list for a multi-scale simulation in which a task at one simulation scale produces an output to be applied as an input to a task at a different simulation scale, the case/task list identifying a plurality of cases to be evaluated, at least one task for each of the cases, and dependencies among the tasks, each of the cases defining a material being evaluated for use in semiconductor technology, and each of the cases further defining a state of the material and differing from all of the other cases in either the material or the material state or both, and the case/task list further identifying, for a subject one of the material cases, at least two tasks of different approaches of the multi-scale simulation including a first approach being an ab initio approach and a second approach being a non-ab initio approach, the case/task list including at least one instance in which a task of the second approach follows a task of the first approach;
• the computer system allocating available processor cores to at least some of the tasks, constrained by the dependencies, and initiating execution of the tasks on allocated cores;
• in response to completion of a particular one of the tasks of a particular one of the cases, the particular task generating a result, making a determination as to whether or not the result indicates that the particular case is not likely to have predetermined characteristics, and if the determination is that the result indicates that the particular case is not likely to have the predetermined characteristics, then terminating or pruning one or more of uncompleted tasks in the case/task list;
• and
• further in response to the completion of the particular task, re-allocating one or more available processor cores to at least one of the pending not-yet-executing tasks in accordance with time required to complete the tasks and constrained by the dependencies, and initiating execution of the tasks on one or more allocated cores.

A computer readable medium having stored thereon in a non-transitory manner, a plurality of software code portions defining logic for optimizing turnaround time for a multi-scale simulation of material cases for use in semiconductor technology, including logic for:

• allocating available processor cores to at least some of the tasks of a case/task list for a multi-scale simulation in which a task at one simulation scale produces an output to be applied as an input to a task at a different simulation scale, the case/task list identifying a plurality of cases to be evaluated, at least one task for each of the cases, and dependencies among the tasks, each of the cases defining a material and a state of the material, each of the materials being for use in semiconductor technology, each of the cases and differing from all of the other cases in either the material or the material state or both, and the case/task list further identifying, for a subject one of the material cases, at least two tasks of different approaches of the multi-scale simulation including a first approach being an ab initio approach and a second approach being a non-ab initio approach, the case/task list including at least one instance in which a task of the second approach follows a task of the first approach;
• initiating execution of the tasks on allocated cores;
• in response to completion of a particular one of the tasks of a particular one of the cases, the particular task generating a result, making a determination as to whether or not the result indicates that the particular case is not likely to have predetermined characteristics, and if the determination is that the result indicates that the particular case is not likely to have the predetermined characteristics, then terminating or pruning one or more of uncompleted tasks in the case/task list;
• and
• further in response to the completion of the particular task, re-allocating one or more available processor cores to at least one of the pending not-yet-executing tasks in accordance with time required to complete the tasks and constrained by the dependencies, and initiating execution of the tasks on one or more allocated cores.

A task control system for managing multi-scale simulations of material cases for use in semiconductor technology, comprising:

• a data processor;
• storage accessible to the processor and including code portions which when executed by the processor cause the processor to perform the steps of:
• receiving a case/task list for a multi-scale simulation in which a task at one simulation scale produces an output to be applied as an input to a task at a different simulation scale, the case/task list identifying a plurality of cases to be evaluated, at least one task for each of the cases, and dependencies among the tasks, each of the cases defining a material and a state of the material,
• the case/task list further identifying, for a subject one of the material cases, at least two tasks of different approaches of the multi-scale simulation including a first approach being an ab initio approach and a second approach being a non-ab initio approach, the case/task list including at least one instance in which a task of the second approach follows a task of the first approach,
• the case/task list including first and second ones of the tasks (e.g. DFT1, DFT2) of the first approach all identifying the same first material but each identifying a different state of the first material, each of the first and second tasks having first and second subtasks (e.g. DFTx.1, DFTx.2) to be executed sequentially;
• allocating available processor cores to at least some of the tasks, constrained by the dependencies, and initiating execution of the tasks on allocated cores, including allocating one or more available processor cores to the first subtask of the first task;
• in response to completion of the first subtask of the first task, initiating execution of the second subtask of the first task on one or more available cores, using intermediate results from the first subtask of the first task, and initiating execution of the second subtask of the second task on one or more available cores, using the intermediate results from the first subtask of the first task (722);
• and
• in response to completion of the first and second tasks, initiating execution of at least one of the pending not-yet-executing tasks which have not been terminated or pruned, using combined results from the first and second tasks.

A method for optimizing turnaround time for a multi-scale simulation of material cases for use in semiconductor technology, comprising the steps of:

• a computer system receiving a case/task list for a multi-scale simulation in which a task at one simulation scale produces an output to be applied as an input to a task at a different simulation scale, the case/task list identifying a plurality of cases to be evaluated, at least one task for each of the cases, and dependencies among the tasks, each of the cases defining a material and a state of the material,
• the case/task list further identifying, for a subject one of the material cases, at least two tasks of different approaches of the multi-scale simulation including a first approach being an ab initio approach and a second approach being a non-ab initio approach, the case/task list including at least one instance in which a task of the second approach follows a task of the first approach,
• the case/task list including first and second ones of the tasks (e.g. DFT1, DFT2) of the first approach all identifying the same first material but each identifying a different state of the first material, each of the first and second tasks having first and second subtasks (e.g. DFTx.1, DFTx.2) to be executed sequentially;
• allocating available processor cores to at least some of the tasks, constrained by the dependencies, and initiating execution of the tasks on allocated cores, including allocating one or more available processor cores to the first subtask of the first task;
• in response to completion of the first subtask of the first task, the computer system initiating execution of the second subtask of the first task on one or more available cores, using intermediate results from the first subtask of the first task, and initiating execution of the second subtask of the second task on one or more available cores, using the intermediate results from the first subtask of the first task (722);
• and
• in response to completion of the first and second tasks, the computer system initiating execution of at least one of the pending not-yet-executing tasks which have not been terminated or pruned, using combined results from the first and second tasks.

A computer readable medium having stored thereon in a non-transitory manner, a plurality of software code portions defining logic for optimizing turnaround time for a multi-scale simulation of material cases for use in semiconductor technology, including logic for:

• receiving a case/task list for a multi-scale simulation in which a task at one simulation scale produces an output to be applied as an input to a task at a different simulation scale, the case/task list identifying a plurality of cases to be evaluated, at least one task for each of the cases, and dependencies among the tasks, each of the cases defining a material and a state of the material,
• the case/task list further identifying, for a subject one of the material cases, at least two tasks of different approaches of the multi-scale simulation including a first approach being an ab initio approach and a second approach being a non-ab initio approach, the case/task list including at least one instance in which a task of the second approach follows a task of the first approach,
• the case/task list including first and second ones of the tasks (e.g. DFT1, DFT2) of the first approach all identifying the same first material but each identifying a different state of the first material, each of the first and second tasks having first and second subtasks (e.g. DFTx.1, DFTx.2) to be executed sequentially;
• allocating available processor cores to at least some of the tasks, constrained by the dependencies, and initiating execution of the tasks on allocated cores, including allocating one or more available processor cores to the first subtask of the first task;
• in response to completion of the first subtask of the first task, initiating execution of the second subtask of the first task on one or more available cores, using intermediate results from the first subtask of the first task, and initiating execution of the second subtask of the second task on one or more available cores, using the intermediate results from the first subtask of the first task (722);
• and
• in response to completion of the first and second tasks, initiating execution of at least one of the pending not-yet-executing tasks which have not been terminated or pruned, using combined results from the first and second tasks.