Science AMA Series: We are Michèle Weiland, Nick Johnson, and Juan Rodriguez Herrera, and we’re members of the Adept project, dedicated to improving energy efficiency in parallel computing. We’ve created a suite of tools designed to do just that. AUA!

See the source article by following the link below:

Hi Reddit!

We are scientists from the Adept Project, a European project investigating energy efficiency in high-performance and Embedded computing. The three of us here today are based at the University of Edinburgh, but the project has several other partners throughout Europe.

The next big goal in high-performance computing is the Exascale – computers capable of a billion billion calculations every second. Computing researchers around the world are aiming to reach this goal by 2022. However, it comes with a cost: energy. Spiralling energy costs, as well as environmental concerns, means that new computers need to be designed to be as energy efficient as possible, whilst still being powerful enough to achieve their goals.

In the Adept project, we have created several different tools, such as the Adept Benchmark Suite, Power Measurement System, and Power and Performance Prediction Tool, to both test real systems and design new ones. You can find out a bit more about each of these on our website, adept-project.eu. Essentially, though, we can use the tools together to predict how a system will use power, how a specific piece of software will perform on a system (even if the system doesn’t exist yet) and design new systems that will perform more efficiently.

Michèle is a Project Manager at EPCC, the supercomputing centre at the University of Edinburgh, and leads the Adept project. She’s interested in research that is trying to work out how to use computers as efficiently as possible, so that we don’t waste energy, power or compute resources!

Nick is a wannabe engineer and researcher. He's been dabbling with computer systems for about 20 years and has spent most of the last 3 trying to measure and understand the trade-offs between runtime and energy consumption in a variety of machines.

Juan joined EPCC as an Applications Developer in December 2015 after completing a PhD in Computer Science at the University of Almeria (Spain). He’s currently involved in several research projects, including Adept, and providing user support to HPC users.

We’ll be here from 9 am ET (3pm BST) to answer your questions about power use, energy efficiency, and high performance and Embedded computing, so Ask Us Anything!

">Science AMA Series: We are Michèle Weiland, Nick Johnson, and Juan Rodriguez Herrera, and we’re members of the Adept project, dedicated to improving energy efficiency in parallel computing. We’ve created a suite of tools designed to do just that. AUA!