| QUESTIONS
|
XPRESS
|
TG X-Stack
|
DEGAS
|
D-TEC
|
DynAX
|
X-TUNE
|
GVR
|
CORVETTE
|
SLEEC
|
PIPER
|
| What performance information would an application developer need to know about software developed in your X-Stack project?
|
|
|
|
|
|
|
|
|
|
|
| What performance information would a code developer in your project need to know about your software?
|
|
|
|
|
|
|
|
|
|
|
| What information should a performance tool gather from each level in your software stack? What performance information should each level of your software stack maintain for inspection by performance tools? How would a tool get this information: register a signal handler, call an async-signal safe function from a signal handler, register a callback, call a function from a registered callback, or something else?
|
|
|
|
|
|
|
|
|
|
|
| What types of performance problems do you want tools to measure and diagnose? Insufficient parallelism? Serialization? Load-imbalance? Task dependences? Data movement costs? Power or energy consumption? Failures and failure handling costs? The overhead of resilience mechanisms (e.g., checkpointing)?
|
|
|
|
|
|
|
|
|
|
|
| What kinds of performance problems do you foresee analyzing using post-mortem analysis?
|
|
|
|
|
|
|
|
|
|
|
| What kinds of performance problems do you foresee analyzing using on-the-fly analysis? What interfaces will be needed to gather the necessary information? What control interfaces will be necessary to enable runtime adaptation based on performance measurements?
|
|
|
|
|
|
|
|
|
|
|
| Current auto-tuning tools treat a program as a black-box. What kind of white-box performance analysis techniques do you envision would be helpful for auto-tuning application performance?
|
|
|
|
|
|
|
|
|
|
|
| There is a gap between the application-level and implementation-level views of programming languages and DSLs. What information should your software layers (compiler and runtime system) provide to attribute implementation-level performance measurement data to an application-level view?
|
|
|
|
|
|
|
|
|
|
|
| What kind of visualization and presentation support for bugs do you want from performance tools? Do you envision any IDE integration for performance tools?
|
|
|
|
|
|
|
|
|
|
|
| List the performance challenges that you think the next generation programming languages and models would face?
|
|
|
|
|
|
|
|
|
|
|
| How do you envision using performance feedback to reduce energy or power consumption? What sort of interfaces will that require for inquiry and control?
|
|
|
|
|
|
|
|
|
|
|
| How can performance tools help with resilience?
|
|
|
|
|
|
|
|
|
|
|