Autonomic SMT System Tuning

Abstract

Methods, systems, and media are disclosed for autonomic system tuning of simultaneous multithreading (“SMT”). In one embodiment, the method for autonomic tuning of at least one SMT setting for an optimized processing, such as via throughput, latency, and power consumption, of a workload on a computer system includes calling, by a kernel, an SMT scheduler having at least one hook into a genetic library. Further, the method includes obtaining, by the SMT scheduler through the at least one hook, genetic data from the genetic library for the optimized processing of the workload. Further still, the method includes tuning, by the SMT scheduler and based on the obtaining, the at least one SMT setting for at least one cpu of the computer system.

1 . A method for autonomic tuning of at least one SMT setting for an optimized processing of a workload on a computer system, the method comprising:
 calling, by a kernel, an SMT scheduler having at least one hook into a genetic library;   obtaining, by the SMT scheduler through the at least one hook, genetic data from the genetic library for the optimized processing of the workload; and   tuning, by the SMT scheduler and based on the obtaining, the at least one SMT setting for at least one cpu of the computer system.   
   
   
       2 . The method of  claim 1 , further comprising running a genetic algorithm to calculate the genetic data. 
   
   
       3 . (canceled) 
   
   
       4 . The method of  claim 2 , further comprising introducing at least one mutation into the running of the genetic algorithm, wherein the at least one mutation is a different setting of the at least one SMT setting for the at least one cpu of the computer system. 
   
   
       5 . The method of  claim 2 , wherein the running comprises continuously running the genetic algorithm. 
   
   
       6 . The method of  claim 2 , wherein the running comprises running the genetic algorithm for a fixed period. 
   
   
       7 . The method of  claim 1 , further comprising storing, by the genetic algorithm, the genetic data in the genetic library. 
   
   
       8 . The method of  claim 1 , wherein the tuning the at least one SMT setting comprises tuning one or more processor threads associated with the at least one cpu of the computer system. 
   
   
       9 . A system for autonomic tuning of at least one SMT setting for an optimized processing of a workload on a computer system, the system comprising:
 a kernel in communication with an SMT scheduler having at least one hook into a genetic library;   a genetic data module of the genetic library, wherein the genetic data module has genetic data necessary for the optimized processing of the workload; and   a control module of the genetic library for providing the genetic data to the SMT scheduler through the at least one hook, whereby the SMT scheduler tunes the at least one SMT setting for at least one cpu of the computer system in accordance with genetic data provided.   
   
   
       10 . The system of  claim 9 , further comprising a learning module, associated with the genetic library, for running a genetic algorithm and calculating the genetic data. 
   
   
       11 . The system of  claim 10 , further comprising a configuration module for configuring the genetic algorithm with an initial value of the at least one SMT setting for the at least one cpu of the computer system, for configuring the genetic algorithm with at least one mutation, and for configuring a metric for the genetic algorithm to calculate the genetic data in order to achieve the optimized processing of the workload. 
   
   
       12 . The system of  claim 9 , wherein the genetic data comprises performance and ranking latency data of the at least one SMT setting for the at least one cpu of the computer system. 
   
   
       13 . The system of  claim 9 , wherein the genetic data comprises performance and ranking throughput data of the at least one SMT setting for the at least one cpu of the computer system. 
   
   
       14 . The system of  claim 9 , wherein the genetic data comprises performance and ranking power consumption data of the at least one SMT setting for the at least one cpu of the computer system. 
   
   
       15 . The system of  claim 9 , wherein the genetic data module further comprises one or more switches for managing the genetic data. 
   
   
       16 . A machine-accessible medium containing instructions, which when executed by a machine, cause the machine to perform operations for autonomic tuning of at least one SMT setting for optimized processing of a workload on a computer system, comprising:
 calling, by a kernel, an SMT scheduler having at least one hook into a genetic library;   obtaining, by the SMT scheduler through the at least one hook, genetic data from the genetic library for the optimized processing of the workload; and   tuning, by the SMT scheduler and based on the obtaining, the at least one SMT setting for at least one cpu of the computer system.   
   
   
       17 . The machine-accessible medium of  claim 16 , wherein the instructions further comprise instructions to perform operations for running a genetic algorithm to generate the genetic data. 
   
   
       18 . The machine-accessible medium of  claim 17 , wherein the instructions further comprise instructions to perform operations for configuring, before performing the running, the genetic algorithm with an initial value of the at least one SMT setting for the at least one cpu of the computer system, and with a metric selected in order to achieve the optimized processing for the workload. 
   
   
       19 . The machine-accessible medium of  claim 17 , wherein the instructions further comprise instructions to perform operations for introducing at least one mutation into the running of the genetic algorithm, wherein the at least one mutation is a different setting of the at least one SMT setting for the at least one cpu of the computer system. 
   
   
       20 . The machine-accessible medium of  claim 17 , wherein the instructions for running comprises instructions for running the genetic algorithm for a fixed period. 
   
   
       21 . The machine-accessible medium of  claim 16 , wherein the instructions further comprise instructions to perform operations for storing, by the genetic algorithm, the genetic data in the genetic library. 
   
   
       22 . The machine-accessible medium of  claim 16 , wherein the instructions for tuning the at least one SMT setting comprises instructions for tuning one or more processor threads associated with the at least one cpu of the computer system.