Cycles per byte

Cycles per byte (sometimes abbreviated cpb) is a unit of measurement which indicates the number of clock cycles a microprocessor will perform per byte (usually of octet size) of data processed in an algorithm.^[1] It is commonly used as a partial indicator of real-world performance in cryptographic functions.^[2]

References

Processing Benchmark

Concepts

Organizations

Processor

Floating point unit (FLOPS)	SPECfp Linpack LAPACK Prime95 Super PI SuperPrime Whetstone (benchmark) IBM iSeries benchmarks (Computational Intensive Workload)
Integer (ALU)	Dhrystone Fhourstones SPECint
Digital signal processor (DSP)	BTDi
Graphics processing unit (GPU)	BRL-CAD
Parallel computing	DEISA Benchmark Suite Livermore loops NAS Parallel Benchmarks HPC Challenge Benchmark Princeton Application Repository for Shared-Memory Computers (PARSEC)

Network	Breaking Point Systems SUPS
Filesystems and Storage	Bonnie++ HD Tach IOzone Diskspd
Computer memory	BSS Random Access benchmark HPC Challenge Random Memory Access
I/O Subsystem	Iometer Ioblazer IBM iSeries benchmarks (Commercial Processing Workload)

Computer system (entire)

Energy consumption

Software

JavaScript engine	Browser speed test
cryptography	Cycles per byte
Multiuser system	SDET AIM Multiuser Benchmark
Virtual machine	VMmark SPECvirt
Physics engine	Physics Abstraction Layer API Bench
Recursion performance	Tak (function)
Database transactions	rPerf (IBM) TATP Benchmark Transaction Processing over XML (TPoX) YCSB (NoSQL)
Web server benchmarking	Apache JMeter Curl-loader httperf OpenSTA TPC-W Tsung
Video Compression	Surveillance performance index
X Window System	Xmark93

Platform specific

This article is issued from Wikipedia. The text is licensed under Creative Commons - Attribution - Sharealike. Additional terms may apply for the media files.