SAP HANA. Jake Klein/ SVP SAP HANA June, 2013

Similar documents
SAP HANA Scalability. SAP HANA Development Team

SAP HANA Update. Saul Cunningham SAP Big Data Centre of Excellence

HANA Performance. Efficient Speed and Scale-out for Real-time BI

In-Memory Data Management Jens Krueger

Oracle Exadata: Strategy and Roadmap

UNFAIR ADVANTAGE Your Road to SAP Hana 2016 PURE STORAGE INC.

SAP HANA as an Accelerator for PLM Processes HANA Basics and Scenarios

Oracle Performance on M5000 with F20 Flash Cache. Benchmark Report September 2011

Introduction to SAP HANA and what you can build on it. Jan 2013 Balaji Krishna Product Management, SAP HANA Platform

Oracle Exadata X7. Uwe Kirchhoff Oracle ACS - Delivery Senior Principal Service Delivery Engineer

Infrastructure Matters: POWER8 vs. Xeon x86

2011 IBM Research Strategic Initiative: Workload Optimized Systems

SAP HANA x IBM POWER8 to empower your business transformation. PETER LEE Distinguished Engineer Systems Hardware, IBM Greater China Group

Introduction to Data Management CSE 344

New Approach to Unstructured Data

A U G U S T 8, S A N T A C L A R A, C A

Copyright 2012 EMC Corporation. All rights reserved.

DDN. DDN Updates. Data DirectNeworks Japan, Inc Shuichi Ihara. DDN Storage 2017 DDN Storage

SAP HANA SAP HANA Introduction Description:

In-Memory Data Management for Enterprise Applications. BigSys 2014, Stuttgart, September 2014 Johannes Wust Hasso Plattner Institute (now with SAP)

1 Copyright 2012, Oracle and/or its affiliates. All rights reserved.

Lenovo Enterprise Portfolio

2 to 4 Intel Xeon Processor E v3 Family CPUs. Up to 12 SFF Disk Drives for Appliance Model. Up to 6 TB of Main Memory (with GB LRDIMMs)

IBM FlashSystem. IBM FLiP Tool Wie viel schneller kann Ihr IBM i Power Server mit IBM FlashSystem 900 / V9000 Storage sein?

SAP High-Performance Analytic Appliance on the Cisco Unified Computing System

VMware Virtual SAN Technology

Sub-Second Response Times with New In-Memory Analytics in MicroStrategy 10. Onur Kahraman

In-Memory Computing EXASOL Evaluation

Approaching the Petabyte Analytic Database: What I learned

Pervasive DataRush TM

Lenovo Database Configuration for Microsoft SQL Server TB

Hitachi Converged Platform for Oracle

Was ist dran an einer spezialisierten Data Warehousing platform?

SAP RDS 에최적화된 IBM Hardware IBM System x3850 X5 and x3690 X5: Workload Optimized Solution for SAP

Optimizing and Modeling SAP Business Analytics for SAP HANA. Iver van de Zand, Business Analytics

Introduction to Database Systems CSE 414

DDN. DDN Updates. DataDirect Neworks Japan, Inc Nobu Hashizume. DDN Storage 2018 DDN Storage 1

VOLTDB + HP VERTICA. page

PSAM, NEC PCIe SSD Appliance for Microsoft SQL Server (Reference Architecture) September 4 th, 2014 NEC Corporation

Recent Innovations in Data Storage Technologies Dr Roger MacNicol Software Architect

Analyze Big Data Faster and Store It Cheaper

IBM DB2 BLU Acceleration vs. SAP HANA vs. Oracle Exadata

The next step in Software-Defined Storage with Virtual SAN

LATEST INTEL TECHNOLOGIES POWER NEW PERFORMANCE LEVELS ON VMWARE VSAN

IBM s Data Warehouse Appliance Offerings

Oracle Exadata. Smart Database Platforms - Dramatic Performance and Cost Advantages. Juan Loaiza Senior Vice President Oracle Database Systems

Lenovo Database Configuration

Evolution of Capabilities Hunter Downey, Solution Advisor

Deep Dive on SimpliVity s OmniStack A Technical Whitepaper

IBM TS4300 with IBM Spectrum Storage - The Perfect Match -

Reconstruyendo una Nube Privada con la Innovadora Hiper-Convergencia Infraestructura Huawei FusionCube Hiper-Convergente

The Impact of SSD Selection on SQL Server Performance. Solution Brief. Understanding the differences in NVMe and SATA SSD throughput

The Arrival of Affordable In-Memory Database Management Systems

In-Memory Data Management

Microsoft Analytics Platform System (APS)

Oracle Server Benchmark with In-Memory SQL Processing Exadata X2-2 half-rack high-capacity

Huawei KunLun Mission Critical Server. KunLun 9008/9016/9032 Technical Specifications

Pervasive Insight. Mission Critical Platform

Introduction to Database Services

FAST SQL SERVER BACKUP AND RESTORE

IBM Power 9 надежная платформа для развертывания облаков. Ташкент. Юрий Кондратенко Cross-Brand Sales Specialist

SAP HANA for Next-Generation Business Applications and Real-Time Analytics

Advances of parallel computing. Kirill Bogachev May 2016

Disclaimer This presentation may contain product features that are currently under development. This overview of new technology represents no commitme

Combine Native SQL Flexibility with SAP HANA Platform Performance and Tools

Modernize Your IT with FlexPod. NetApp & Schneider Electric

Copyright 2012 EMC Corporation. All rights reserved.

Design a Remote-Office or Branch-Office Data Center with Cisco UCS Mini

SAP HANA ADMINISTRATION

Looking ahead with IBM i. 10+ year roadmap

How Might Recently Formed System Interconnect Consortia Affect PM? Doug Voigt, SNIA TC

Flash In the Data Center

Pivot3 Acuity with Microsoft SQL Server Reference Architecture

Leading Performance for Oracle Applications? John McAbel Collaborate 2015

Fast Hardware For AI

Huawei KunLun Mission Critical Server. KunLun 9008/9016/9032 Technical Specifications

DataON and Intel Select Hyper-Converged Infrastructure (HCI) Maximizes IOPS Performance for Windows Server Software-Defined Storage

Costefficient Storage with Dataprotection

Oracle #1 for Data Warehousing. Data Warehouses Growing Rapidly Tripling In Size Every Two Years

Toward a Memory-centric Architecture

SAP HANA Inspirience Day

Understanding the SAP HANA Difference. Amit Satoor, SAP Data Management

HP visoko-performantna OLTP rješenja

What Really Sets Apart Exadata from the Rest

Aerospike Scales with Google Cloud Platform

Achieving Memory Level Performance: Secrets Beyond Shared Flash

INTEL NEXT GENERATION TECHNOLOGY - POWERING NEW PERFORMANCE LEVELS

How to Deploy Enterprise Analytics Applications With SAP BW and SAP HANA

Persistent Memory. High Speed and Low Latency. White Paper M-WP006

Accelerating Digital Transformation with InterSystems IRIS and vsan

Harnessing the Potential of SAP HANA with IBM Power Systems

Netezza The Analytics Appliance

SQL Server Everything built-in

2016 IBM Corporation 1

NEW CONVERGED APPROACH FOR SAP POWERED BY ATOS

Harnessing the potential of SAP HANA with IBM Power Systems

SAP NetWeaver BW Performance on IBM i: Comparing SAP BW Aggregates, IBM i DB2 MQTs and SAP BW Accelerator

Secure, scalable storage made simple. OEM Storage Portfolio

The Future of High Performance Computing

Building NVLink for Developers

Transcription:

SAP HANA Jake Klein/ SVP SAP HANA June, 2013

SAP 3 YEARS AGO Middleware BI / Analytics Core ERP + Suite

2013 WHERE ARE WE NOW? Cloud Mobile Applications SAP HANA Analytics D&T

Changed Reality Disruptive Performance 4 CPU w/ 10 cores each (optional 8CPU -> 80cores) 512GB RAM (for 4CPU) 1TB RAM (for 8CPU) 2012 SAP AG. All rights reserved. 6

Programming for a New Platform CPU Core CPU Cache Main Memory Disk Performance bottleneck today: CPU waiting for data to be loaded from memory into cache Performance bottleneck in the past: Disk I/O 2012 SAP AG. All rights reserved. 7

L1 cache ~ 1ns 64k 2012 SAP AG. All rights reserved. 10

L2 cache ~ 5ns 256k 2012 SAP AG. All rights reserved. 11

L3 cache ~ 20ns 8M 2012 SAP AG. All rights reserved. 12

Main memory ~ 100ns TBs 2012 SAP AG. All rights reserved. 13

Disk > 1.000.000ns PBs Source: Google Maps 2012 SAP AG. All rights reserved. 14

Price per GByte Moore s Law: DRAM Pricing 1000x 100x 10x 1 TB ~ $10K Multi-Terabyte servers are now completely affordable! Source: OBJECTIVE ANALYSIS http://blogs-images.forbes.com/jimhandy/files/2011/12/dram-gb-price.jpg 2012 SAP AG. All rights reserved. 15

Moore s Law: CPUs 2002 2007 2013 2014 1 core 32 bits 4MB 2 cores 2 CPUs per server External Controllers 8 Sockets X 10 Cores 160 threads per server On-chip memory control Quick interconnect VM and vector support 64 bits; 256 GB - 1 TB More cores, bigger caches 480 Threads per server Greater on-chip integration (PCIe, network,...) Data-direct I/O Tens - hundreds of TBs 2012 SAP AG. All rights reserved. Images: Intel, Danilo Rizzuti / FreeDigitalPhotos.net 16

In-memory computing SAP HANA HW technology innovations Multicore architecture (8 x 10 core CPU per blade) Massive parallel scaling with many blades One blade ~$50,000 = 1 enterprise class server SAP SW technology innovations Row and column store Compression Partitioning 64 bit address space 2 TB in current servers 100 GB/s data throughput Dramatic decline in price/performance No aggregate tables Insert only on delta Convergence of improved hardware economics and technology innovations enables SAP to deliver on its vision of the real-time enterprise with in-memory business applications 2012 SAP AG. All rights reserved. 17

Operational Reporting Oil and Gas We use about 140,000 different materials, which creates several million records per year which have to be analyzed to determine which materials should go to which locations. The possibility to work directly with operational data in a real time mode with large volumes of our data is a tremendous opportunity. CIO, Oil and Gas Company 16 hours before HANA 2 minutes with HANA (HANA processing time 3 sec) 2012 SAP AG. All rights reserved. 18

ERP Accelerator - CPG Prior to SAP HANA, we were unable to run full analytics in a reasonable timeframe. With SAP HANA, we will be able to run analytics at a local level on specific brands and locations, and at the lowest level of detail in real time. CIO, CPG Company 4 hours before SAP HANA few seconds with SAP HANA 2012 SAP AG. All rights reserved. 19

Suite On HANA Customer Experience SAP CO/PA Accelerator ERP w/o HANA ERP with HANA Acceleration factor vs. ERP EBIT with commodity sales initial report EBIT with commodity sales drilldown by alphacode Cost allocation initial report Cost allocation drilldown by sending cost center 280 sec 7 sec (DB 2,8 sec)* 620 sec 5 sec (DB 2,9 sec)* 45 sec 5 sec (DB 3,4 sec)* 260 sec 7 sec (DB 3,3 sec)* 40 124 9 37 Additional drilldowns are now possible e.g. by customer and article (not feasible before) Drilldown performance not dependent on availability of suitable aggregation level, selection always on line item level from In-Memory Database Existing ERP reports are accelerated with no changes to report definitions *DB measurements show the selection runtime on HANA. Non-DB time is likely to decrease if the application server runs on production hardware. 2012 SAP AG. All rights reserved. 20

THE HANA EFFECT 1500+ Customers 30 in 10K Club +700 on HANA ONE 25 Verticals 69 Solutions 483 startups 9 HW Vendors Transactions Analytics NLP Text Search & Analysis GIS PAL XS Hadoop 2B /sec /core 10M /sec /core 1M /sec

SAP HANA Software Component View SQL SQL Script MDX Other Analytical and special interfaces Text Analytics Application Function Libraries Business Function Library Predictive Analysis Library Application logic extensions Parallel Calculation engine Parallel data flow computing model Relational Stores Row based Columnar Object Graph Store Multiple in-memory stores Managed Appliance Appliance packaging 2012 SAP AG. All rights reserved. 22

The Five Dimensions of HANA Applications Dimension Complex Questions Large, Multi-type data Recent Data Interactive Full, Raw Data Expectations of Technology Programmatic ability to go Deep, apply complex analytics, predictive and statistical techniques Ability to go Broad, combine structured and text search on massive amounts of data Real Time Computing and response High Performance No pre-fabrication, aggregation, tuning If you need three out of five it is an excellent fit for HANA If you need all five - only HANA can support you!! 2012 SAP AG. All rights reserved. 24

SAP HANA Global Partner Eco System SAP HANA Hardware & Technology Partners 2013 SAP AG. All rights reserved. 27

Thank you