Java Performance Tuning
Generation Scavenging: A non-disruptive high performance storage reclamation algorithm
SDE 1 Proceedings of the first ACM SIGSOFT/SIGPLAN software engineering symposium on Practical software development environments
An Information Exploration Tool for Performance Analysis of Java Programs
TOOLS '01 Proceedings of the Technology of Object-Oriented Languages and Systems
Better, Faster, Lighter Java
The causes of bloat, the limits of health
Proceedings of the 22nd annual ACM SIGPLAN conference on Object-oriented programming systems and applications
Validity of the single processor approach to achieving large scale computing capabilities
AFIPS '67 (Spring) Proceedings of the April 18-20, 1967, spring joint computer conference
Communications of the ACM - Security in the Browser
Hi-index | 0.00 |
Enterprise software shows increasing levels of concurrency and complexity and decreasing "think times" between user interactions. Such trends are evident in both emerging workloads, such as social networking, and traditional applications, such as banking, for which both query counts and complexity are increasing. Similarly, in today's multicore-processor era, processor core counts double every processor generation. However, not all hardware capacity (e.g., cache, disk, and network capacity) is growing at this core rate. As a result, processor dies will have more cores sharing resources. Personnel associated with our project Multicore Applications Restructured for Scaling started with a well-tuned baseline version of a large multitier commercial workload and worked to eff iciently identify a small set of software changes that, together, would lead to improved scaling and performance. This paper reports the required tooling and analysis, software changes, and improvements (factor performance gains). Perhaps most surprising, we found that a small set of code changes allowed major performance gains within a single Java®virtual machine, which has implications for the way in which we develop and deploy software.