Since some years energy efficiency in datacenters is improving but the amount of electricity needed for operating those for hosting Cloud services is growing with the sizes of the infrastructure and the user demands. With renewable energies and the usage of direct current in datacenters, we believe that we can cope with this problem. On the one hand several efforts have been conducted at the computing level in datacenters (for service placement and scheduling) and, on the other hand, for energy provisioning partly with renewable energies, but without much interaction between these two efforts. DATAZERO will give consistent solutions for high availability of IT (Information Technologies) services, avoiding unnecessary redundancies, under the constraints of intermittent nature of electrical and services flows. We noticed that, despite a true interest towards pragmatic and market-based solutions, comprehensive studies are missing to cope with the robustness of a cloud infrastructure in the conditions of datacenters powered by renewable energy. The question we address in DATAZERO is: How to manage the electricity and the service flows in order to deliver services to customers in a robust and efficient manner within datacenters operated with several energy sources? Any solution for electricity and IT management must maintain the health state of the infrastructure in nominal conditions. IT services must be resilient to electrical problems and their operation must be assured. The methodology we propose is: 1. using the characterization of power sources, we will optimize the decision process managing how to balance electricity production and storage among potential power sources (multisource), how to optimize the electricity flow in the system, in order to ensure a given level of constrained energy demand. 2. using the characterization of hosted services in the cloud, we will optimize the decision process solving how to balance the service execution among the computers, how to schedule and place hosted services in the system, in order to ensure resilience in a dynamic and prone-to-failure power distribution system. 3. using the two preceding decision processes, we will introduce a negotiation loop able to match the constraints coming from both sides (adapting energy production depending on the computing demand and adapting computing service level depending on the energy offer) and optimize the renewable energy utilization. While the computer level and the energy level can be treated separately, we believe there is a need to formulate a global optimization problem mixing constraints, scheduling and requests from both sides. IT tasks dispatching should take into account the source capacities and, depending on the foreseen workflow, the different energy sources should be used at their optimal functioning. At energy level, hybridization of different sources (or several sources with lower power than the maximum requested) is challenging in the architecture design (dimensioning, type of suitable sources…) and global management of energy flow transiting between sources is of paramount importance. The main targets are middle-size datacenters where IT load can be managed either through Virtualization or Cloud orchestrator (up to 1000 m2 and 1 MW) commonly seen in enterprise and public institutions. The main outcomes are the optimization of datacenters powered partially with renewable sources and the production of a simulation toolkit. When developed this simulation toolkit will allow for testing, tuning and comparing several mixes of renewable sources, electrical and computer equipment, and scheduling policies. It will provide recommendations on the IT and electrical redundancies in hybrid architectures for reaching a given level of performance. DATAZERO brings together an interdisciplinary team (academic and industry, from both IT and electrical engineering) to find suitable solutions in a cooperative way.

Over the last decade, several companies and universities have investigated the possibilities of using renewable energies to decrease the large amount (around 200 TWh) of brown energy classically used by datacenters. Pioneering companies have managed to fuel their datacenters with renewable energies (Eolas in France, Emerson Network Power, AISO, Apple, Google, Facebook, Amazon... more recently), but either the green energy accounts only for a share of the total energy (datacenters consumption of green energy represents about 3 GWh) and the energy production is actually outsourced to a different site (i.e., they inject the same amount of power that they take from the power grid), or they buy Renewable Energy Certificates System (RECS) certificates from an electricity provider producing renewable energy to claim that their datacenter is green. All rely additionally on at least one external electricity provider for power redundancy. Few research initiatives as FP7 GreenDataNet, involving partner EATON, and ANR DATAZERO, involving all partners of DATAZERO2, have paved the way to study how a datacenter mainly operated by renewable energies works. Since 2015 and till 2019, the ANR DATAZERO project investigates new ways to power and manage one Cloud data center using ONLY on-site renewable energy sources and storage equipment (photovoltaics, wind turbines, hydrogen-energy, batteries, supercaps). However DATAZERO lacks a careful integration of uncertainty in the decision processes, relies on classical power and cooling infrastructures and considers only one datacenter. As a first objective, DATAZERO2 aims at improving the operation and the design of datacenters ONLY operated with renewable energies, in particular using disruptive approaches for uncertainty management, cooling and power distribution, and considering multiple datacenters. A second objective is to propose a knowledge base and toolbox for industry and researchers for zero-emission datacenters beyond state-of-the-art solutions. From a scientific point of view, the following challenges will be addressed: Challenge 1 - Uncertainty mitigation. Challenge 2 - Reconfigurable power distribution system. Challenge 3 - Datacenter cooling from hydrogen technologies. Challenge 4 - Trading energy for computation and vice-versa. Challenge 5 - Conducting Life Cycle Analysis of renewable energy powered datacenters. In addition to these research challenges, we will carry on a technical challenge: Challenge 6 - Producing an open source testbed for understanding renewable energy powered datacenters. DATAZERO2 builds on the same hypotheses that were the strengths of DATAZERO: not relying on a Power Grid and focus on small to midsize datacenters up to 1 MW. DATAZERO2 is a multidisciplinary project gathering experts from computer science, electricity, energy and automation: UPS-IRIT (Computer Science), FEMTO-ST with three departments (DISC for Computer Science, AS2M for Automation, and ENERGIE for Energy and Electrical Engineering), INPT-LAPLACE (Electrical Engineering), and EATON SAS, an industrial partner specialised in power provisioning in datacenters and links with computing infrastructures. It builds over valuable expertise from the ANR DATAZERO project, helping to accelerate the understanding the complexity of the problems at hand.

MANGO targets to achieve extreme resource efficiency in future QoS-sensitive HPC through ambitious cross-boundary architecture exploration for performance/power/predictability (PPP) based on the definition of new-generation high-performance, power-efficient, heterogeneous architectures with native mechanisms for isolation and quality-of-service, and an innovative two-phase passive cooling system. Its disruptive approach will involve many interrelated mechanisms at various architectural levels, including heterogeneous computing cores, memory architectures, interconnects, run-time resource management, power monitoring and cooling, to the programming models. The system architecture will be inherently heterogeneous as an enabler for efficiency and application-based customization, where general-purpose compute nodes (GN) are intertwined with heterogeneous acceleration nodes (HN), linked by an across-boundary homogeneous interconnect. It will provide guarantees for predictability, bandwidth and latency for the whole HN node infrastructure, allowing dynamic adaptation to applications. MANGO will develop a toolset for PPP and explore holistic pro-active thermal and power management for energy optimization including chip, board and rack cooling levels, creating a hitherto inexistent link between HW and SW effects at all layers. Project will build an effective large-scale emulation platform. The architecture will be validated through noticeable examples of application with QoS and high-performance requirements. Ultimately, the combined interplay of the multi-level innovative solutions brought by MANGO will result in a new positioning in the PPP space, ensuring sustainable performance as high as 100 PFLOPS for the realistic levels of power consumption (<15MWatt) delivered to QoS-sensitive applications in large-scale capacity computing scenarios providing essential building blocks at the architectural level enabling the full realization of the ETP4HPC strategic research agenda

By 2020, more than 7 billion people and businesses, and close to 20.8 billion devices, will be connected to the Internet. To host these data, datacenters (DC) have become the backbone of the digital society as well as of a large part of our economy and their number and complexity keeps increasing rapidly. Businesses of all sizes are moving their key processes to external colocation datacenters. “The global datacenter colocation size is expected to grow from USD 31.52 billion in 2017 to USD 62.30 billion by 2022, at a Compound Annual Growth Rate (CAGR) of 14.60%.” At the same time, datacenter resources require monitoring and management to ensure proper business continuity. To that extent, IDEAS project will enable datacenters to benefit significantly from the value add of DCIM (datacenter Infrastructure Management) solutions at a fraction of the cost and complexity of today. DCIM features will be accessible through a service-oriented business model. These services will address the needs present during the active life of datacenters and enforce growth of datacenter industry and associated industries across the EU. An example of this approach is the remote monitoring of the datacenter infrastructure using the intuitive / immersive interface of IDEAS. It will provide an innovative and competitive answer at a time when the number of datacenters is dramatically increasing in Europe due to colocation, “cloudification” of enterprise IT, but also new trends such as IoT, 5G that will result in the surge of very small and scattered datacenters (also known as Edge IT). The proposed project is needed to unlock the significant potential of commercialization of R&D results (novel datacenter monitoring software) developed in the ongoing FET project MANGO. It will use this newly developed know-how and existing 3D engines to bring this system not only to a limited research world but also to a large number of datacenter operators.