Appendix:Glossary of energy efficient hardware/software

The list of terms and abbreviations from the field of energy efficient hardware/software, used within the H2020 M2DC project, has been compiled to unify terminology. Conceived as a live document to be updated with new terms.

is an IT automation and orchestration framework, designed to automate routine tasks like server/application deployment and configuration. (Application Programming Interface) is a set of defined methods used for communication between various (SW) modules. (AutoRegressive Integrated Moving Average) is a statistical model for time series data to analyze its behavior or predict future values (Advanced RISC Machine) describes a family of architectures for computer processors building on reduced instruction set computing (RISC). M2DC makes use of CPUs based on the latest ARMv8 64bit architecture (ETH, PCI, PWR, MGMT) the four different PCBs used for connections in the back of the chassis. Variants for Ethernet, PCI Express, power connections between chassis, "management" connections (Management Ethernet port, HDMI, USB, ...). is the PCB to which the baseboards are connected. Includes a management controller. One of the backplanes in a system will also act as a "chassis controller" (running the RECS_Master). describes the process of deploying and running operating systems and applications directly on physical hardware without further abstraction layers like hypervisors. In M2DC context, bare-metal provisioning/deployment usually includes mechanisms for dynamic and/or automatic deployment. is the PCB to which the microservers are mounted. Includes a management controller. Available as high performance (COM Express) and two low power (Jetson, Apalis) versions. is a software tool for gathering and presentation of time-series monitoring data like power, temperatures, CPU/GPU/memory/network utilization etc. (Baseboard Management Controller) is a microcontroller embedded on the server mainboard, whose task is to manage the interface between system management software and hardware as part of IPMI. is a statistical mean to stabilize variance of (time series) data. The mechanical module that the user plugs into the chassis. Contains either one (high performance) or two (low power) baseboards. "Carrier blade" – the user's point of view while "Baseboard" is the physical (PCBs) and logical (In the management software) view. Cache coherency SEE is the monitoring component of OpenStack collecting and saving data about physical and virtual resources as well as triggering actions if certain conditions are met. is an OpenStack Nova extension to support better scaling mechanisms by partitioning compute nodes into smaller groups coupled each with their own database and queue instead of using one single database and queue. The whole M2DC server chassis. Will be available as small chassis, mid-range chassis and scale-out chassis. Convolutional neural networks SEE is a network-based Linux installation server Component Object Model Remote DMA and MPI communication SEE Cyber-Physical Systems, computing resources combined with sensors/actuators Central Processing Unit Symmetric and homomorphic encryption SEE is a monitoring service for cloud infrastructures and applications (Data Centre Infrastructure Management) describes a set of tools (especially software) which help to manage the datacentres infrastructure. It organizes information coming from different components (e.g. servers, temperatures, air conditioners) and is often able to turn components on and off. (Data Center Manager) see DCIM Dataflow runtime SEE Data mining SEE provides an additional layer of system virtualization abstraction and automation. By using resource isolation, the overhead of Docker containers is much smaller than VMs Ecommerce: CloudStore use case HPC numerical weather forecast with EULAG use case (Field Programmable Gate Array) is an integrated circuit which can be configured after manufacturing as it contains an array of programmable logic blocks and a hierarchy of reconfigurable interconnects. PCB in the front of the chassis. Provides 2x USB, 1x HDMI, power button, small joystick and a 128x128 display. is a time series database as a service and may extend the \gls{Ceilometer} component for efficient data saving (Graphics Processing Unit) is an electronic circuit specialized on processing large blocks of data in parallel and therefore efficient for manipulating computer graphics and image processing. (Generic Resource) is a resource (other than CPU/nodes) which can be added to SLURM (High Performance Computing) is a field of computer-aided data processing with high computation and/or storage demand.
 * Ansible
 * API
 * ARIMA
 * ARM
 * Backpanel
 * Backplane
 * Bare-metal provisioning
 * Baseboard
 * BEMOS
 * BMC
 * Box-Cox
 * Carrier blade
 * CC SEE
 * Ceilometer (Telemetry)
 * Cells
 * Chassis
 * CNN SEE
 * Cobbler
 * COM
 * COM SEE
 * CPS
 * CPU
 * CRYPT SEE
 * DataDog
 * DCIM
 * DCM
 * DM SEE
 * Docker
 * ECOM use case
 * EULAG use case
 * FPGA
 * Frontpanel
 * Gnocchi
 * GPU
 * GRES
 * HPC

High performance data analytics with R use case (Holt-Winters filtering) is a method for exponential smoothing of time series data. enable the usage of virtualization on physical machines by managing HW demand for all VMs running on the host (Infrastructure as a Service) is a standardized, highly automated offering, where compute resources, complemented by storage and networking capabilities are owned and hosted by a service provider and offered to customers on-demand. (Internet Control Message Protocol) supports the exchange of information and error messages via the internet protocol (IP). Every router and computer is supposed to support ICMP. Intrusion detection SEE (Intrusion Detection System) is any device or software application that monitors a network or systems aiming to detect malicious activity or policy violations. (Intrusion Prevention System) is a network security/threat prevention technology that examines network traffic flows to detect and prevent vulnerability exploits. (Intelligent Management) represents the interface component connecting the management components with data sources, models and control mechanisms (Nova, hardware) in the context of M2DC Image processing SEE Low cost image processing use case (Internet of Things) is the network of physical devices, vehicles, home appliances, and other items embedded with electronics, software, sensors, actuators, and network connectivity which enable these objects to connect and exchange data. IoT data analytics for transportation use case (Internet Protocol) is a broadly applied network protocol, which is also the base for the Internet. IP delivers data packets from source to destination hosts based on IP addresses in packet headers (Intelligent Platform Management Bus) connects the BMC with other management controllers in the system (satellite controllers) (Intelligent Platform Management Interface) is a standard interface for computer hardware and firmware. It supports to control and monitor computer hardware remotely, even if they are turned off. is the OpenStack component which enables dynamic bare-metal provisioning analogue virtualization, but with less overhead (JavaScript Object Notation) is a compact and easy to read text format to exchange data between applications (Local Area Network) is a computer network interconnecting nodes within a limited space such as residential or office buildings is the server platform developed within the M2DC project, which will be sold by Christmann under the selling name RECS®|Box. is a programming framework for processing big data sets with a parallel algorithm on distributed compute nodes Microcontroller Unit (Many Integrated Core) describes a CPU architecture from Intel, which combines many CPU cores on one chip The COM Express/Jetson/Apalis servers doing the actual computational work. "Microserver" as the main term while "node" can be used as a synonym to not over-use "microserver". Preprocessing monitoring data SEE Message passing SEE Multi-core Processing Unit, a processing system composed of two or more independent cores is a monitoring software for complex IT infrastructures (networks, hosts and services). It is extendable by different modules (Compute) is the component in the OpenStack cloud manager responsible for provisioning compute instances. Nova also integrates its own scheduler to select target nodes based on different parameters (Nagios Remote Plugin Executor) is a monitoring client for Nagios running on hosts. It communicates to the Nagios server (Topology and Orchestration Specification for Cloud Applications) is a standard language to describe a topology of cloud based web services, their components, relationships, and processes developed by non-profit organization OASIS describes a software or system, which runs on company owned or rented hardware, rather than in an external cloud is a popular cloud management software, which is used as base platform for middleware development in the M2DC project (Platform as a service) is a category of cloud computing services that provides a platform allowing customers to develop, run, and manage applications without the complexity of building and maintaining the infrastructure Energy optimized PaaS Printed Circuit Board Peripheral Component Interconnect (Peripheral Component Interconnect Express) is a high speed bus standard with hot-plug functionality (PHP: Hypertext Preprocessor) is a scripting language which is especially used for dynamic web pages and web applications (Power Management) executes management decisions from RTM and WM like switching nodes on/off either directly or via Nova API is the term for polling plugins in Ceilometer PCB to which the power supplies connect. Contains a small microcontroller to manage standby/power-on of the chassis. is an open-source administration software for automated computer configurations via network (Preboot eXecution Environment) extends the booting process of computers by automatically fetching the appropriate operating system image via LAN is an open source enterprise ready assets management focused DCIM and equipment lifecycle management system is the selling name of microserver chassis sold by Christmann. The fourth revision of RECS|Box is the M2DC server, which is the server platform developed within the M2DC project. can be installed on compute modules in RECS|Box systems. It communicates to the RECS|Box management and supports the OS-level monitoring by sending data, while it is also able to execute commands received from the management is the embedded controller in the RECS|Box system which is also responsible for monitoring and serves as an interface to the hardware is an open industry standard specification and schema to support the management of scalable platform hardware (REpresentational State Transfer) is a definition for implementing web interfaces REST offering a uniform and predefined set of stateless operations for requesting systems (Remote Management Control Protocol) is a protocol used for communication to and from the BMC (Resource and Thermal Management) is responsible for managing the thermal and power environment (Software as a Service) is part of the cloud computing model. IT infrastructure SaaS and software is centrally hosted and software usage is licensed on a subscription basis (Seasonal AutoRegressive Integrated Moving Average) is the ARIMA model extended with an additional component to model seasonal data (Software Development Kit) is a collection of tools and libraries to support/ease software development (Sensor Data Record) is a data set with information about one specific sensor. Such data sets are stored in a central, non-volatile memory, called SDR repository (System Efficiency Enhancements) describes the M2DC concept of combined hardware/software modules which can extend appliances to increase efficiency in various aspects (energy/power, throughput, ...) (Search Engine Optimization) is the process of affecting the online visibility of a website or a web page in a web search engine's unpaid results. (Simple Linux Utility for Resource Management) is a job scheduler for Linux/Unix used in the high performance and super computing field (Simple Network Management Protocol) is a standard network protocol to SNMP monitor and control network devices (e.g. servers, routers, switches). It describes the network package structure and the communication procedure (System on a Chip) means the integration of all or most functions of an electronic system on one chip (System on Module) is a board-level circuit that integrates a system function in a single module.
 * HPDA use case HW
 * HW
 * Hypervisor
 * IaaS
 * ICMP
 * ID SEE
 * IDS
 * IPS
 * IM
 * IMG SEE
 * IMG use case
 * IoT
 * IoT use case
 * IP
 * IPMB
 * IPMI
 * Ironic
 * JSON
 * LAN
 * M2DC server
 * MapReduce
 * MCU
 * MIC
 * Microserver/Node
 * MON SEE
 * MP SEE
 * MPU
 * Nagios
 * Nova
 * NRPE
 * OASIS TOSCA
 * On-premises
 * OpenStack
 * PaaS
 * PAAS
 * PCB
 * PCI
 * PCIe
 * PHP
 * PM
 * Pollster
 * Power distribution board
 * Puppet
 * PXE
 * Ralph
 * RECS®|Box
 * RECSDaemon
 * RECS_Master
 * Redfish API
 * REST
 * RMCP
 * RTM
 * SaaS
 * SARIMA
 * SDK
 * SDR
 * SEE
 * SEO
 * SLURM
 * SNMP
 * SoC
 * SoM