2017 IEEE International Symposium on Workload Characterization

Characterization of applications in domains such as

• Cloud, grid, search engines, e-commerce, web, databases, file/application servers

• Embedded, real-time, Edge Computing, mobile and extensible distributed systems, multimedia

• Memory, storage, and file systems

• Life sciences, bioinformatics, scientific computing, finance, forecasting

• Management and/or analytics big data, machine Learning, Analytics, Data mining and other data-intensive computation

• Security, reliability, biometrics

• User behavior and system-user interaction, social computing

• Cyber-physical systems, pervasive computation and Internet of Things (IoT)

Characterization of workload on emerging technology, such as

• Quantum computation and communication

• Near-threshold Computation

• Non-volatile memory

Characterization of OS, Virtual Machine, middleware and library behavior

• Virtual machines, Websphere, .NET, Java VM, databases

• Graphics libraries, scientific libraries

• Operating system and hypervisor effects and overheads

Implications of workloads in design issues, such as

• Reliability, performance, power, security and privacy, scalability, and sustainability

• Processors, memory hierarchy, I/O, and networks

• Performance accelerators, FPGA’s, GPU’s, etc.

• Near data processing architectures

• Novel architectures (non-Von-Neumann)

Benchmark creation and evaluation, including

• Multithreaded benchmarks, benchmark cloning

• Profiling, trace collection, synthetic traces

• Validation of benchmarks

Measurement tools and techniques, including

• Instrumentation methodologies for workload verification and characterization

• Techniques for accurate analysis/measurement of production systems

• Analytical and abstract modeling of program behavior and systems