Gabriel Aul: N-State Chaos and Windows Performance

Download this episode

Download Video

Description

Gabriel Aul is a 16 year Microsoft veteran who has always had his hand in the chaotic world of system performance and reliability, from technical product support and testing to being a leader on the Windows performance team. In fact, he was part of the team that developed the original Watson failure reporting tool. He's a dev at heart (once a dev, always a dev) and understands the complexities of Windows as a platform: Windows supports thousands of devices and the thousands of drivers that make them useful to users (devices sometimes have more than one driver, so add that to the complexity quotient...). How can so many devices (drivers) work together successfully (meaning not hosing the system) with so many supported configurations and possibilities for drivers to bring Windows to a screeching halt (think about task scheduling, resource allocation, background processing, foreground processing user mode code execution, kernel mode code execution and the sheer amount of concurrent running code, all over the place, all contending for Windows' attention...).

The world of Windows as platform is incredibly complex (n-state chaotic). It's amazing, actually, that the chaos doesn't lead to more performance and reliability issues. The Windows performance team has some really powerful tools (and an effort called Velocity) that can help ISVs find highly complex performance issues. Gabriel touches on this in this conversation and we address the continued need for more guidance and samples for developers.

One thing is for sure: The developers who write applications and drivers for Windows are great developers. You are craftsman, artists. Windows is a great platform in many ways because of the people who innovate on top of it. You know who you are. Yeah, you. Thank you.

Embed

Format

Available formats for this video:

Actual format may change based on video formats available and browser capability.

    The Discussion

    Comments closed

    Comments have been closed since this content was published more than 30 days ago, but if you'd like to continue the conversation, please create a new thread in our Forums, or Contact Us and let us know.