BRL-CAD is a powerful constructive solid geometry solid modeling system that includes an interactive geometry editor, ray-tracing support for rendering and geometric analysis, path-tracing for realistic image synthesis, network distributed framebuffer support, and image and signal-processing tools.
PTT helps users to analyze and understand correction and performance problems for multi-threaded applications. It shows when a program calls NPTL routines and when it exits from them, with details about the internal mechanisms of the library. It is a post mortem analysis; the trace can be analyzed once the application has stopped. Three different trace formats are provided: a machine readable text format, a human readable text format, and a graphical format. The graphical format is obtained thanks to Pajé, an interactive visualization tool.
Performance Co-Pilot (PCP) is a framework and set of services for supporting system-level performance monitoring and performance management. It provides a unifying abstraction for all of the interesting performance data in a system, and allows client applications to easily retrieve and process any subset of that data using a single API. A client-server architecture allows multiple clients to monitor the same host, and a single client to monitor multiple hosts. Archive logging and replay are integrated so that a client application can use the same API to process real-time data from a host or historical data from an archive.
RUBiS is an auction site modeled after eBay.com used to benchmark e-commerce Web site technologies. It is currently used to evaluate design patterns, application servers, and communication layers scalability. Several implementations using PHP, Servlets, Enterprise JavaBeans (EB BMP, EB CMP, MDB, SB, EJB 2.0 CMP, Session Façade, etc.) are already available and new versions for JDO and .Net are currently developed.
kernbench is a CPU throughput benchmark. It is designed to compare kernels on the same machine, or to compare hardware. It runs a kernel compile at various numbers of concurrent jobs: 1/2 number of CPUs, optimal (default is 4xnumber of CPUs), and maximal job count. Optionally it can also run single threaded. It then prints out a number of useful statistics for the average of each group of runs.