Accelerating Datacentre performance through Memory Chips to efficiently manage the Big Data Age

Project Information

ENERGIA/PIM

Grant agreement ID: 190141232

DOI 10.3030/190141232

Project closed

EC signature date 22 April 2022

Start date 1 March 2022

End date 29 February 2024

Funded under

• The European Innovation Council (EIC)

Total cost € 3 566 042,50

EU contribution € 2 496 229,75

Coordinated by

SAS UPMEM
France

Project description

Faster and more efficient datacentres

Datacentres underpin all digital activity across all sectors. Without them, everything would stop. However, they consume 2 % of worldwide electricity and this will rise to 10 % by 2030. Most of the energy (80 %) is spent on moving data between the main memory chips and the processor (CPU). There is a way to change this. The development of Processing-In-Memory (PIM) technology prevents this data movement by performing calculations in the memory chips, where the data resides. No longer necessary to move data off chips, PIM bypasses structural bottlenecks. The EU-funded ENERGIA/PIM project will facilitate the uptake of this technology. Top selling points include the fact that PIM-equipped servers are up to 20 times faster and 10 times more efficient.

Objective

Datacentres have become a backbone of the modern economy, they currently consume 2% of worldwide electricity, rising to 10% by 2030. While compute accounts for 40% of this energy, 80% of this compute energy is related to moving data between the main memory chips and the processor (CPU), a structural bottleneck within server design that causes slowdown and low CPU usage. Development of ground-breaking Processing-In-Memory (PIM) technology by UPMEM prevents this data movement by performing calculations in the memory chips, where the data resides. By eliminating the need to move data off chips, PIM bypasses structural bottlenecks. The power of UPMEM’s technology comes from leveraging existing industry protocols, chip technology and programing languages, and designing PIM modules that fit into standard memory slots. When compared to those using conventional
memory, PIM equipped servers have proven to be up to 20x faster, 10x more efficient and cheaper for Big Data and AI applications.

Fields of science

• engineering and technology › electrical engineering, electronic engineering, information engineering › electronic engineering › computer hardware › computer processors
• natural sciences › computer and information sciences › data science › big data

Keywords

• Memory chips
• Processing in memory
• Data centres
• Big Data
• Severs

Programme(s)

• HORIZON.3.1 - The European Innovation Council (EIC) Main Programme
• HORIZON.3.1.2 - The Accelerator

Topic(s)

• HORIZON-EIC-2021-ACCELERATORCHALLENGES-01-01 - Strategic Digital and Health Technologies

Call for proposal

HORIZON-EIC-2021-ACCELERATORCHALLENGES-01

Funding Scheme

HORIZON-EIC-ACC-BF - HORIZON EIC Accelerator Blended Finance

Coordinator

SAS UPMEM

• Net EU contribution: € 2 496 229,75
• Address: 6 PLACE ROBERT SCHUMAN ; 38000 Grenoble ; France
• SME The organization defined itself as SME (small and medium-sized enterprise) at the time the Grant Agreement was signed. Yes:
  Region
  Auvergne-Rhône-Alpes › Rhône-Alpes › Isère
• Region: Auvergne-Rhône-Alpes › Rhône-Alpes › Isère
• Activity type: Private for-profit entities (excluding Higher or Secondary Education Establishments)
• Total cost: € 3 566 042,50