An EUREKA Celtic-plus project SE NDAT E - E XT E ND Adj. prof. Tor Björn Minde, CEO RISE SICS North AB (Head of Strategy Ericsson Research)
SE NDAT E 2 5,4 M€ 73 M€
A ne w industr y e r a and tr ansfor mation • Digitalization – traditional industries • Data – great amount, owned & shared • Things – connected and intelligent • Communications – faster and more • Users – active and enhanced • Functions – more and seamless The digitalization looks like the industrialization
HPC vs. We b sc ale - IT HPC Mainframe Web-scale IT (High Performance Computing) (Hyper scale computing) Compute intensive Data intensive (Simulations, calculations) (Big Data analysis, Media) Data moved to Algorithms moved to pieces of Algorithms pieces of Data Scale-up Architecture Scale-out Architecture
Me ga datac e nte r c omple xity Source: Nyteknik 2012
E XT E ND – Holistic Datac e nte r Automation Holistic Datacenter Automation for efficient cross layer optimization
E XT E ND – Holistic Datac e nte r Automation Applic a tion Ha rdwa re Ne tworking Softwa re Assumptions Cooling Building Assumptions Assumptions Assumptions Assumptions Assumptions Powe r IT IT Contr ol Contr ol Contr ol Contr ol Contr ol Contr ol Contr ol 20.11.2017 7
E XT E ND – Holistic Datac e nte r Automation Common Polic ie s, Shar e d Data(base ), Common Optimization Cr ite r ia Applic a tion Ha rdwa re Ne tworking Softwa re Cooling Building Powe r Knowle dge Knowle dge Knowle dge Knowle dge Knowle dge Knowle dge IT IT Contr ol Contr ol Contr ol Contr ol Contr ol Contr ol Contr ol 20.11.2017 8
E XT E ND – Common Me tr ic s Data(base ) and Data Colle c tion Environmental Sensors MODBUS etc. Datacenter Actuation and control Analytics tools Streaming platform SNMP etc. OPC-UA Data collection and short term Long term storage storage Metrics collection 20.11.2017 9
SICS- ICE Datac e nte r http://ic ice.s .sics ics.s .se In operation since January 2016. 5 TB of metrics data now
Root Ca use Ana lytic s ba se d on me tr ic c or r e la tion Source Ericsson/SICS
Pr oje c t par tne r s Large enterprises Medium enterprises Small enterprises Academia Financial support 20.11.2017 14
supported by
Recommend
More recommend