Gil Tene, CTO and Co-founder, Azul Systems

Gil Tene is CTO and co-founder of Azul Systems. He has been involved with virtual machine technologies for the past 20 years and has been building Java technology-based products since 1995. Gil pioneered Azul's Continuously Concurrent Compacting Collector (C4), Java Virtualization, Elastic Memory, and various managed runtime and systems stack technologies that combine to deliver the industry's most scalable and robust Java platforms.

In 2006 he was named one of the Top 50 Agenda Setters in the technology industry by Silicon.com. Prior to co-founding Azul, Gil held key technology positions at Nortel Networks, Shasta Networks and at Check Point Software Technologies, where he delivered several industry-leading traffic management solutions including the industry's first Firewall-1 based security appliance. He architected operating systems for Stratus Computer, clustering solutions at Qualix/Legato, and served as an officer in the Israeli Navy Computer R and D unit. Gil holds a BSEE from The Technion Israel Institute of Technology, and has been awarded 27 patents in computer-related technologies.

Twitter: @giltene

Presentation: "How not to measure latency"

Track: Solutions Track # 1

Time: Wednesday 13:30 - 14:20

Location: Regency

Abstract:

Understanding application responsiveness and latency is critical to delivering good application behavior. But good characterization of bad data is useless. When measurements of response time present false or misleading latency information, even the best analysis can lead to wrong operational decisions and poor application experience.

In this talk, Gil Tene (CTO, Azul Systems) will discuss some common pitfalls encountered in measuring and characterizing latency. Gil will demonstrate and discuss some false assumptions and measurement techniques that lead to incorrect results, and cover simple ways to sanity check and correct these situations. He will discuss the fallacy of using standard deviation measurements, the strongly multi-modal nature of latency, common discontinuities found in most computing platforms, and how back pressure and coordinated data omission issues can skew measurement results dramatically. Gill will introduce and demonstrate how simple and recently open sourced tooling can be used to improve and gain higher confidence in both latency measurement and reporting.

Presentation: "JVM mechanics: a peak under the hood."

Track: Mechanical Sympathy

Time: Friday 11:40 - 12:30

Location: Seacliff CD

Abstract:

The Java Virtual Machine is just that - a machine. The ways in which this machine actually performs the work you ask of it is often quite interesting, and sometimes surprising. Gaining an intuition into what executing your code actually involves can help you write better, higher performing code. Even more importantly, understanding what is NOT involved in executing your code can save days of wasted premature optimization work. In this talk, Gil Tene (CTO, Azul Systems, and a JVM mechanic) will discuss examples of how the freedom this machine has in re-interpreting the meaning of code can have dramatic implications on performance and other code behavior. Gil will show examples of speculative, observation-based runtime optimizations that can often dominate the actual behavior of many code paths. He will demonstrate the importance of ordering and memory model rules in controlling optimizations, and discuss some of the basics of how the garbage collection "machine" functions, and the key elements controlling both efficiency and responsiveness in memory management. We will conclude with interactive Q&A on JVM mechanics.