Multi Mu ti-Di Dimensio mensional nal, In-GP GPU-Memor Memory - - PowerPoint PPT Presentation

multi mu ti di dimensio mensional nal in gp gpu memor
SMART_READER_LITE
LIVE PREVIEW

Multi Mu ti-Di Dimensio mensional nal, In-GP GPU-Memor Memory - - PowerPoint PPT Presentation

Multi Mu ti-Di Dimensio mensional nal, In-GP GPU-Memor Memory Dat ataba abases ses: St Stre ream aming ing Co Condi dition tional al Ca Calculat culations ions in in Bi Big g Data a Se Sets Peter Strohm | GTC15 | San


slide-1
SLIDE 1

Bus usiness Driven Intellig lligence, Analytics & Performance Management. www.jedox.com - @JedoxAG - @PSJedox - #gtc15 - #s5481

Mu Multi ti-Di Dimensio mensional nal, In-GP GPU-Memor Memory Dat ataba abases ses:

St Stre ream aming ing Co Condi dition tional al Ca Calculat culations ions in in Bi Big g Data a Se Sets

Peter Strohm | GTC15 | San José | 03/17/2015

slide-2
SLIDE 2

Bus usiness Driven Intellig lligence, Analytics & Performance Management. www.jedox.com - @JedoxAG - @PSJedox - #gtc15 - #s5481

2002 Founded in Freiburg, Germany Today Offices in Freiburg, Frankfurt, Düsseldorf, Paris, Boston 120 Global Business Partners 125 Countries with Jedox Users 25 Language Versions

2

Jedox

SA App

slide-3
SLIDE 3

Bus usiness Driven Intellig lligence, Analytics & Performance Management. www.jedox.com - @JedoxAG - @PSJedox - #gtc15 - #s5481

Big Data in real-time: Twitter

slide-4
SLIDE 4

Bus usiness Driven Intellig lligence, Analytics & Performance Management. www.jedox.com - @JedoxAG - @PSJedox - #gtc15 - #s5481

Big Data in real-time: Twitter

slide-5
SLIDE 5

Bus usiness Driven Intellig lligence, Analytics & Performance Management. www.jedox.com - @JedoxAG - @PSJedox - #gtc15 - #s5481

Twitter: Big Data in Real-time

slide-6
SLIDE 6

Bus usiness Driven Intellig lligence, Analytics & Performance Management. www.jedox.com - @JedoxAG - @PSJedox - #gtc15 - #s5481

In-Memory OLAP-Database

All data in main memory 1

Jan Feb Mar Q1 Apr May Jun Jul Q2 Aug Sep Q3 Oct Nov Dec Q4 Year

All regions Europe France Italy UK North America USA Canada Mexico

Count TweetID Longitude

In-GPU-Memory OLAP-Database

Jan Feb Mar Apr May Jun Jul Aug Sep Dec Nov Oct

Q1 Q2 Q3 Q4 Year

Hierarchical structure of consolidated elements 2 All data in GPU memory

slide-7
SLIDE 7

Bus usiness Driven Intellig lligence, Analytics & Performance Management. www.jedox.com - @JedoxAG - @PSJedox - #gtc15 - #s5481

In-Memory OLAP-Database

In-GPU-Memory & “on-the-fly” i

In-GPU-Memory OLAP-Database

Store only non-zero values 4 Calculation “on-the-fly” 3

Jan Feb Mar Q1 Apr May Jun Jul Q2 Aug Sep Q3 Oct Nov Dec Q4 Year

All regions Europe France Italy UK North America USA Canada Mexico

Count TweetID Longitude

slide-8
SLIDE 8

Bus usiness Driven Intellig lligence, Analytics & Performance Management. www.jedox.com - @JedoxAG - @PSJedox - #gtc15 - #s5481

Persistent Twitter Data in OLAP

add

Update trending with Web+ETL

update

Adding Twitter data in ETL

Jedox Social Analytics Workflow

filtering

Streaming via php in . Web

Public Twitter Stream

streaming analyzing

www

Mobile SA App Web ExcelPLUS

slide-9
SLIDE 9

Bus usiness Driven Intellig lligence, Analytics & Performance Management. www.jedox.com - @JedoxAG - @PSJedox - #gtc15 - #s5481

Jedox Social Analytics with GPU

System: 2 x K40 with 12 GB GPURAM CPU with 128GB RAM 1 Data sets with > 500 million entries 2 Calculating “on-the-fly” on al all data entries 3

slide-10
SLIDE 10

Bus usiness Driven Intellig lligence, Analytics & Performance Management. www.jedox.com - @JedoxAG - @PSJedox - #gtc15 - #s5481

Jedox Social Analytics Showcase

slide-11
SLIDE 11

Bus usiness Driven Intellig lligence, Analytics & Performance Management. www.jedox.com - @JedoxAG - @PSJedox - #gtc15 - #s5481

Jedox Social Analytics Showcase

slide-12
SLIDE 12

Bus usiness Driven Intellig lligence, Analytics & Performance Management. www.jedox.com - @JedoxAG - @PSJedox - #gtc15 - #s5481

Social Media Analytics: HeatMap

  • ver 500 Million database entries
  • ver 250.000 grid points
slide-13
SLIDE 13

Bus usiness Driven Intellig lligence, Analytics & Performance Management. www.jedox.com - @JedoxAG - @PSJedox - #gtc15 - #s5481

Social Media Analytics: Trending

Top 100 Trending Factor 1 ~ 28.000 dictionary words > 80 Million DB entries 2 # current day # all previous days ________________

slide-14
SLIDE 14

Bus usiness Driven Intellig lligence, Analytics & Performance Management. www.jedox.com - @JedoxAG - @PSJedox - #gtc15 - #s5481

Social Media Analytics: Trending

1 Assigning Trending Factor to vi virtual cell cells s (e.g. [TF] = [B] / [C]) 2 Trending Factor is calc lculated on

  • n the fly

fly 3 Conditional Calculation: IF IF-Rule le IF IF ([count] >1000) TH THEN [count,currentDay] / [count,previousWeek] ELS ELSE 0

slide-15
SLIDE 15

Bus usiness Driven Intellig lligence, Analytics & Performance Management. www.jedox.com - @JedoxAG - @PSJedox - #gtc15 - #s5481

Conditional Rule with constants

Rule Processor > Processor Constant (0)

[TF] = IF ([B] > 1000) THEN [C] ELSE 0

Matching Processor Constant (1000) Matching Processor

slide-16
SLIDE 16

Bus usiness Driven Intellig lligence, Analytics & Performance Management. www.jedox.com - @JedoxAG - @PSJedox - #gtc15 - #s5481

Conditional Rule with constants

Rule Processor Constant (0)

[TF] = IF ([B] > 1000) THEN [C] ELSE 0

Matching Processor

>

1 1 1 … …

slide-17
SLIDE 17

Bus usiness Driven Intellig lligence, Analytics & Performance Management. www.jedox.com - @JedoxAG - @PSJedox - #gtc15 - #s5481

Conditional Rule with constants

Rule Processor Constant (0)

[TF] = IF ([B] > 1000) THEN [C] ELSE 0

>

1 1 1 … …

T

A,1 A,2 A,3 A,4 A,5 … …

slide-18
SLIDE 18

Bus usiness Driven Intellig lligence, Analytics & Performance Management. www.jedox.com - @JedoxAG - @PSJedox - #gtc15 - #s5481

Conditional Rules with constants

Rule Processor

[TF] = IF ([B] > 1000) THEN [C] ELSE 0

>

1 1 1 … …

T

A,1 A,2 A,3 A,4 A,5 … …

T

slide-19
SLIDE 19

Bus usiness Driven Intellig lligence, Analytics & Performance Management. www.jedox.com - @JedoxAG - @PSJedox - #gtc15 - #s5481

Conditional Rules on GPU

Rule Processor

[TF] = IF ([B] > 1000) THEN [C] ELSE 0

>

1 1 1 … …

T

A,1 A,2 A,3 A,6 A,8 … …

FALSE TRUE

T

slide-20
SLIDE 20

Bus usiness Driven Intellig lligence, Analytics & Performance Management. www.jedox.com - @JedoxAG - @PSJedox - #gtc15 - #s5481

Conditional Rules on GPU

Rule Processor

[TF] = IF ([B] > 1000) THEN [C] ELSE 0

>

1 1 1 … …

T

A,1 A,2 A,3 A,6 A,8 … …

FALSE TRUE

T

slide-21
SLIDE 21

Bus usiness Driven Intellig lligence, Analytics & Performance Management. www.jedox.com - @JedoxAG - @PSJedox - #gtc15 - #s5481

Conditional Rules on GPU

Rule Processor

[TF] = IF ([B] > 1000) THEN [C] ELSE 0

>

1 1 1 … …

T

A,1 A,2 A,3 A,6 A,8 … …

FALSE TRUE complete set

T

slide-22
SLIDE 22

Bus usiness Driven Intellig lligence, Analytics & Performance Management. www.jedox.com - @JedoxAG - @PSJedox - #gtc15 - #s5481

Conditional Rules on GPU

GPU MEMORY SPACE FOR CALCULATIONS

slide-23
SLIDE 23

Bus usiness Driven Intellig lligence, Analytics & Performance Management. www.jedox.com - @JedoxAG - @PSJedox - #gtc15 - #s5481

Conditional Rule on GPU

Rule Processor > Processor Constant (1)

[TF] = IF ([B] > 1000) THEN [C] ELSE 1

Matching Processor Constant (1000) Matching Processor

Streaming Streaming Streaming

slide-24
SLIDE 24

Bus usiness Driven Intellig lligence, Analytics & Performance Management. www.jedox.com - @JedoxAG - @PSJedox - #gtc15 - #s5481

Streaming in Conditional Rules

Rule Processor > Processor Matching Processor

[TF] = IF ([B] > 1000) THEN [C] ELSE [D]

Matching Processor Constant (1000) Matching Processor

Part 1 Part 2 Part 1 Part 2 Total

slide-25
SLIDE 25

Bus usiness Driven Intellig lligence, Analytics & Performance Management. www.jedox.com - @JedoxAG - @PSJedox - #gtc15 - #s5481

Streaming in Conditional Rules

Rule Processor > Processor Matching Processor

[TF] = IF ([B] > 1000) THEN [C] ELSE [D]

Matching Processor Constant (1000) Matching Processor

Part 1 Part 2 Part 1 Part 2 Total

FALSE TRUE

slide-26
SLIDE 26

Bus usiness Driven Intellig lligence, Analytics & Performance Management. www.jedox.com - @JedoxAG - @PSJedox - #gtc15 - #s5481

Streaming in Conditional Rules

Rule Processor > Processor Matching Processor

[TF] = IF ([B] > 1000) THEN [C] ELSE [D]

Matching Processor Constant (1000) Matching Processor

Part 2

Part 1

Part 2

Sub- Total

slide-27
SLIDE 27

Bus usiness Driven Intellig lligence, Analytics & Performance Management. www.jedox.com - @JedoxAG - @PSJedox - #gtc15 - #s5481

Streaming in Conditional Rules

Rule Processor > Processor Matching Processor

[TF] = IF ([B] > 1000) THEN [C] ELSE [D]

Matching Processor Constant (1000) Matching Processor

Part 2 Part 2 1

slide-28
SLIDE 28

Bus usiness Driven Intellig lligence, Analytics & Performance Management. www.jedox.com - @JedoxAG - @PSJedox - #gtc15 - #s5481

Streaming in Conditional Rules

Rule Processor > Processor Matching Processor

[TF] = IF ([B] > 1000) THEN [C] ELSE [D]

Matching Processor Constant (1000) Matching Processor

Part 2 Part 2 Total 1

slide-29
SLIDE 29

Bus usiness Driven Intellig lligence, Analytics & Performance Management. www.jedox.com - @JedoxAG - @PSJedox - #gtc15 - #s5481

Streaming in Conditional Rules

Rule Processor > Processor Matching Processor

[TF] = IF ([B] > 1000) THEN [C] ELSE [D]

Matching Processor Constant (1000) Matching Processor

Part 2 Part 2 Total 1

FALSE TRUE

slide-30
SLIDE 30

Bus usiness Driven Intellig lligence, Analytics & Performance Management. www.jedox.com - @JedoxAG - @PSJedox - #gtc15 - #s5481

Streaming in Conditional Rules

Rule Processor > Processor Matching Processor

[TF] = IF ([B] > 1000) THEN [C] ELSE [D]

Matching Processor Constant (1000) Matching Processor

1

Part 2

Sub- Total

slide-31
SLIDE 31

Bus usiness Driven Intellig lligence, Analytics & Performance Management. www.jedox.com - @JedoxAG - @PSJedox - #gtc15 - #s5481

Streaming in Conditional Rules

Rule Processor > Processor Matching Processor

[TF] = IF ([B] > 1000) THEN [C] ELSE [D]

Matching Processor Constant (1000) Matching Processor

1 2

slide-32
SLIDE 32

Bus usiness Driven Intellig lligence, Analytics & Performance Management. www.jedox.com - @JedoxAG - @PSJedox - #gtc15 - #s5481

Streaming in Conditional Rules

Rule Processor

[TF] = IF ([B] > 1000) THEN [C] ELSE [D]

complete result set

> Processor Matching Processor Matching Processor Constant (1000) Matching Processor

slide-33
SLIDE 33

Bus usiness Driven Intellig lligence, Analytics & Performance Management. www.jedox.com - @JedoxAG - @PSJedox - #gtc15 - #s5481

Features: Performance with GPU

Top 100 1 Heat Map 2 Trending Factor 3 Speed-up up to 50x and more! ! ! In-GPU-Memory database & calculations

slide-34
SLIDE 34

Bus usiness Driven Intellig lligence, Analytics & Performance Management. www.jedox.com - @JedoxAG - @PSJedox - #gtc15 - #s5481

Jedox Social Analytics Showcase

Jedox Social Analytics App download at: www.jedox.com/en/jedox-social-analytics GPU- & Social Analytics App-Team: Alexander Haberstroh Steffen Wittmer Leo Mehlig

slide-35
SLIDE 35

Bus usiness Driven Intellig lligence, Analytics & Performance Management. www.jedox.com - @JedoxAG - @PSJedox - #gtc15 - #s5481

THANK YOU