Prod Pr oduc uctio tion Int Intel elligen igence ce - - PowerPoint PPT Presentation

▶
prod pr oduc uctio tion int intel elligen igence ce
SMART_READER_LITE
LIVE PREVIEW

Prod Pr oduc uctio tion Int Intel elligen igence ce - - PowerPoint PPT Presentation

Sep 29, 2022 285 likes •497 views

Prod Pr oduc uctio tion Int Intel elligen igence ce GPU-Databases GP atabases for or Pre redict ctive ive Maint ntenance enance and d In-Line In Line Con ontr trolling olling in Autom omob obile le Manu nufactu facturi

slide-1
SLIDE 1

Bus usiness Driven Intelli lligence & & Performance Man anagement www.jedox.com – @JedoxAG – @PSJedox – peter.strohm@jedox.com

Pr Prod

  • duc

uctio tion Int Intel elligen igence ce

GP GPU-Databases atabases for

  • r Pre

redict ctive ive Maint ntenance enance and d In In-Line Line Con

  • ntr

trolling

  • lling in Autom
  • mob
  • bile

le Manu nufactu facturi ring ng

slide-2
SLIDE 2

Bus usiness Driven Intelli lligence & & Performance Man anagement www.jedox.com – @JedoxAG – @PSJedox – peter.strohm@jedox.com

8 Facts about Jedox AG

2002

Founded in Freiburg Germany

1

Unified platform for BI & PM

27

Language versions

127

Countries with Jedox users

140,000+

Happy users

1,900+

Organizations using Jedox

8

Locations worldwide

180

Business partners globally

2

slide-3
SLIDE 3

Bus usiness Driven Intelli lligence & & Performance Man anagement www.jedox.com – @JedoxAG – @PSJedox – peter.strohm@jedox.com 3

Jedox - In-Memory OLAP database

slide-4
SLIDE 4

Bus usiness Driven Intelli lligence & & Performance Man anagement www.jedox.com – @JedoxAG – @PSJedox – peter.strohm@jedox.com Source: VDI

slide-5
SLIDE 5

Bus usiness Driven Intelli lligence & & Performance Man anagement www.jedox.com – @JedoxAG – @PSJedox – peter.strohm@jedox.com

slide-6
SLIDE 6

Bus usiness Driven Intelli lligence & & Performance Man anagement www.jedox.com – @JedoxAG – @PSJedox – peter.strohm@jedox.com

Production Intelligence

Measurement systems Automobile supplier Analytics & Databases Inline-Measurement techniques Funded by

slide-7
SLIDE 7

Bus usiness Driven Intelli lligence & & Performance Man anagement www.jedox.com – @JedoxAG – @PSJedox – peter.strohm@jedox.com

Production Intelligence

Local analysis Predictive Maintenance

slide-8
SLIDE 8

Bus usiness Driven Intelli lligence & & Performance Man anagement www.jedox.com – @JedoxAG – @PSJedox – peter.strohm@jedox.com

Massive data input in Real-time

Production line

Additional data

slide-9
SLIDE 9

Bus usiness Driven Intelli lligence & & Performance Man anagement www.jedox.com – @JedoxAG – @PSJedox – peter.strohm@jedox.com

Jedox OLAP data structure

slide-10
SLIDE 10

Bus usiness Driven Intelli lligence & & Performance Man anagement www.jedox.com – @JedoxAG – @PSJedox – peter.strohm@jedox.com

Bulk-Insertion with GPU

slide-11
SLIDE 11

Bus usiness Driven Intelli lligence & & Performance Man anagement www.jedox.com – @JedoxAG – @PSJedox – peter.strohm@jedox.com

Bulk-Insertion: Performance

slide-12
SLIDE 12

Bus usiness Driven Intelli lligence & & Performance Man anagement www.jedox.com – @JedoxAG – @PSJedox – peter.strohm@jedox.com

Production Intelligence

Local analysis Predictive Maintenance

slide-13
SLIDE 13

Bus usiness Driven Intelli lligence & & Performance Man anagement www.jedox.com – @JedoxAG – @PSJedox – peter.strohm@jedox.com

iO redo niO iO redo niO iO redo niO iO redo niO iO redo niO iO redo niO iO redo niO iO redo niO iO redo niO iO redo niO iO redo niO iO redo niO iO redo niO iO redo niO iO redo niO iO redo niO iO redo niO iO redo niO iO redo niO iO redo niO iO redo niO iO redo niO iO redo niO iO redo niO iO redo niO iO redo niO iO redo niO iO redo niO

Predictive Maintenance

A B C D E F G

iO redo niO iO redo niO iO redo niO iO redo niO iO redo niO iO redo niO iO redo niO

slide-14
SLIDE 14

Bus usiness Driven Intelli lligence & & Performance Man anagement www.jedox.com – @JedoxAG – @PSJedox – peter.strohm@jedox.com

DNN in OLAP GPU database

cuDNN

slide-15
SLIDE 15

Bus usiness Driven Intelli lligence & & Performance Man anagement www.jedox.com – @JedoxAG – @PSJedox – peter.strohm@jedox.com

DNN in OLAP GPU database

2 1 Working directly on the GPU database memory

Different input classes by database aggregation/rules

A Time sequence given by database time dimension B (Un-)Supervised learning and feedback by default

slide-16
SLIDE 16

Bus usiness Driven Intelli lligence & & Performance Man anagement www.jedox.com – @JedoxAG – @PSJedox – peter.strohm@jedox.com

DNNs: Current work

2 1 Long Short Time Memory

Unsupervised Learning

3 Backpropagation with database

slide-17
SLIDE 17

Bus usiness Driven Intelli lligence & & Performance Man anagement www.jedox.com – @JedoxAG – @PSJedox – peter.strohm@jedox.com

Industry 4.0 with Jedox + GPU

2 1 GPU as enabling technology

Big data input + analysis

3 New market opportunities

cuDNN

4 DNN for mid-sized companies

slide-18
SLIDE 18

Bus usiness Driven Intelli lligence & & Performance Man anagement www.jedox.com – @JedoxAG – @PSJedox – peter.strohm@jedox.com

Jedox Self-Service BI Suite

Jedox Social Analytics App download at: www.jedox-social-analytics.com 1 Jedox Base Version (with ExcelPLUS) download at: www.jedox.com 2 Webinars & Tutorials at: www.jedox.com/en/webinars ww.jedox.com/en/discover 3

slide-19
SLIDE 19

Bus usiness Driven Intelli lligence & & Performance Man anagement www.jedox.com – @JedoxAG – @PSJedox – peter.strohm@jedox.com

THANK YOU