Skip to Content

SAP Standard Application Benchmarks

Benchmark Procedure APO Cloud Concurrent CRM ERP Industries Netweaver Power Archive
Benchmark Procedure APO Cloud Concurrent CRM ERP Industries Netweaver Power Archive

SAP Standard Application Benchmarks

SAP Standard Application Benchmarks help customers and partners find the appropriate hardware configuration for their IT solutions. Working in concert, SAP and our hardware partners developed the SAP Standard Application Benchmarks to test the hardware and database performance of SAP applications and components.

Overview
Procedure
Benefits
Measuring in SAPS
Sizing
Performance

Advanced Mixed Load (BW-AML)

The SAP BW Advance Mixed Load Benchmark (BW-AML Benchmark) is the successor of the popular BW-EML benchmark which was retired recently. The new BW-AML meets the most current demands of typical business warehouse customers.

Main differences/extensions between BW-EML and BW-AML:

Benchmark Operation

The old BW-EML Benchmark was designed for optimizing query throughput only. The BW-AML now introduces a second phase which aims at response time optimization, with strong focus of customer relevant use cases / scenarios.

Data Model
  • Inserts / Updates / Deletes on transactional data during delta loads
  • Master data changes along with delta loads
  • Utilization of navigation attributes
  • Multiple currencies in source data
  • Stand-alone data loads to measure data load times (Phase 2)
  • Master data lookups, non-trivial transformations (Phase 2)
Queries
  • Randomization of selection of key figures, changing set of key figures in navigation step
  • More sophisticated use of FEMS queries and exception aggregation
  • Currency conversion
  • Extension of all queries and navigation steps on multiple part providers
  • Different query patterns for Info
  • Cube / DSO multiproviders

The BW-AML benchmark consists of 2 phases. Phase 1 is for measuring query throughput at a defined number of initial records. Phase 2 collects information about the runtimes of complex queries and complex delta loads.

The BW-AML benchmark KPIs are for Benchmark Phase 1 the query throughput measured in Advanced Query Navigation Steps / hour. For Phase 2 the 10 complex query runtimes result in the KPI Normalized mean runtime single query test. Finally, the delta load of phase 2 loading, transforming and aggregating 50 Million records results in the Total runtime delta load/transformation test KPI.

Important: Due to the changes in the data model, query design and KPIs the BW-EML and BW-AML benchmark results are not comparable.

Data-mart (BI-D)

The data mart scenario is one use of the Business Intelligence capabilities of SAP NetWeaver. The data mart contains a static snapshot of operational data. Multiple users run queries on this data in 10 InfoCubes which contain 2,500,000,000 records. The key figure is the number of query navigation steps/hour.

Procedure

The following actions are simulated in an SD InfoCube.

  1. . Query 1: Select a year
  2. . Within the year, drill down to a specific country
  3. . Within the country, drill down to a specific sales organization
  4. . Within the sales organization, drill down to distribution channel
  5. . Within the distribution channel, drill down to type, version, and material number
  6. . Change material number and customer number
  7. . Start new query including formulas
  8. . Query 2: Start a new query that retrieves the yearly and quarterly sales volumes for different types of ranges, e.g. sales organizations
  9. . Query 3: Similar to query 1 for country, sales organization, and distribution channel
  10. . Within this query, expand to material level two
  11. . Expand to level three

SAP BI Data Mart Standard Application - Two-Tier: SAP NetWeaver 2004s

Date of Certification (mm/dd/yyyy) Technology Partner Query Navigation Steps/ Hour Operating System - Release RDBMS Release SAP NetWeaver  Release Central Server Number of Processors Number of Cores Number of Threads CPU Type CPU Speed (MHz) CPU Cache Central Server Memory (MB) Certification Number
3/30/2010 Fujitsu 854649 SuSE Linux Enterprise Server 10 Oracle 11g Real Application Clusters (RAC) SAP NetWeaver 7.0 (non-Unicode) Fujitsu PRIMERGY RX600 S5 4 32 64 Intel Xeon Processor X7560 2260 4 KB L1 cache and 256 KB L2 cache per core, 24 MB L3 cache per processor 131072 2010011
10/12/2009 Fujitsu 320363 SuSE Linux Enterprise Server 10 Oracle 10g Real Application Clusters (RAC) SAP NetWeaver 7.0 (non-Unicode) Fujitsu Primergy RX300-S5 2 8 16 Intel Xeon Processor X5570 2930 64 KB L1 cache and 256 KB L2 cache per core, 8 MB L3 cache per processor 98304 2009036
10/31/2008 IBM 182112 i 6.1 DB2 for
i 6.1		 SAP NetWeaver 7.0 (2004s) IBM Power System 570 4 8 16 POWER6 5000 128 KB L1 cache and 4 MB L2 cache per core, 32 MB L3 cache per processor 131072 2008063
4/8/2008 IBM 40445 i 6.1 DB2 for
i 6.1		 SAP NetWeaver 7.0 (2004s) IBM Power System 520 Express 1 2 4 POWER6 4200 128 KB L1 cache and 4 MB L2 cache per core 32768 2008020
10/22/2007 IBM 26224 i5/OS V5R4 DB2 for i5/OS V5R4 SAP NetWeaver 7.0 (2004s) IBM System i 520 2 2 4 POWER5+ 1900 32 KB(D) + 64 KB(I) L1 cache per processor, 1.9 MB L2 cache and 36 MB L3 cache per 2 processors 16384 2007061
7/24/2007 IBM 92716 i5/OS V5R4M5 DB2 for i5/OS V5R4M5 SAP NetWeaver 7.0 (2004s) IBM System i 570 2 4 8 POWER6 4700 128 KB L1 cache and 4 MB L2 cache per core, 32 MB L3 cache per processor 65536 2007047
4/19/2007 IBM 114687 i5/OS V5R4 DB2 for i5/OS V5R4 SAP NetWeaver 2004s IBM System i5 570 8 8 16 POWER5+ 2200 32 KB(D) + 64 KB(I) L1 cache, 1.9 MB L2 cache, 36 MB L3 cache 131072 2007027
01/25/2007 IBM 51875 i5/OS V5R4 DB2 for i5/OS V5R4 SAP NetWeaver 2004s IBM System i5 570 4 4 8 POWER5+ 2200 32 KB(D) + 64 KB(I) L1 cache, 1.9 MB L2 cache, 36 MB L3 cache 65536 2007003

Enlarge
View table

SAP BI Data Mart Standard Application - Three-Tier: SAP NetWeaver 7.0

Date of Certification (mm/dd/yyyy) Technology Partner Query Navigation Steps/ Hour Operating System – Release RDBMS Release SAP NetWeaver  Release Database Servers Database Server Memory (MB) Number & Type of Application Servers Certification Number
5/31/2010 Fujitsu 1624629 SUSE Linux Enterprise Server 10 Oracle 11g Real Application Clusters (RAC) SAP NetWeaver 7.0 (non-Unicode) 2 x Fujitsu Primergy RX600 S5, 4 processors / 32 cores / 64 threads, Intel Xeon Processor X7560, 2,26 GHz, 64 KB L1 cache and 256 KB L2 cache per core, 24 MB L3 cache per processor; 1 server 128 and one server 256 GB main memory 1 x 131072 and 1 x 262144 ran on database server 2010024
11/9/2009 Fujitsu 1,165,742 SuSE Linux Enterprise Server 10 Oracle 10g Real Application Clusters (RAC) SAP NetWeaver 7.0 (non-Unicode) 4 x Fujitsu Primergy RX300-S5, 2 processors / 8 cores / 16 threads, Intel Xeon Processor X5570, 2.93 GHz, 64 KB L1 cache and 256 KB L2 cache per core, 8 MB L3 cache per processor 4 x 98304 ran on database server 2009045
11/9/2009 Fujitsu 900.309 SuSE Linux Enterprise Server 10 Oracle 10g Real Application Clusters (RAC) SAP NetWeaver 7.0 (non-Unicode) 3 x Fujitsu Primergy RX300-S5, 2 processors / 8 cores / 16 threads, Intel Xeon Processor X5570, 2.93 GHz, 64 KB L1 cache and 256 KB L2 cache per core, 8 MB L3 cache per processor 3 x 98304 ran on database server 2009044
10/12/2009 Fujitsu 609349 SuSE Linux Enterprise Server 10 Oracle 10g Real Application Clusters (RAC) SAP NetWeaver 7.0 (non-Unicode) 2 x Fujitsu Primergy RX300-S5, 2 processors / 8 cores / 16 threads, Intel Xeon Processor X5570, 2.93 GHz, 64 KB L1 cache and 256 KB L2 cache per core, 8 MB L3 cache per processor 2 x 98304 ran on database server 2009037

 


Enlarge
View table

Mixed Load (BI-XML)

The mixed load scenario is one use of the Business Intelligence capabilities of SAP NetWeaver.

In this scenario, query activity and load/update activity are executed in parallel. Multiple users run queries on data in 10 SD InfoCubes. You can choose from the following size categories for the initial data load:

  • 300 million records
  • 1000 million records
  • 3000 million records

In three phases during high load activity, static operational data is extended with delta data.
The key figure of this benchmark is the number of query navigation steps/hour.
The BI mixed load benchmark can be run with or without BI Accelerator.

Procedure

The following actions are simulated in an SD InfoCube. Three times during the benchmark run (total 1 hour), a delta data package is uploaded. The total size of these packages is 1/1000th of the total records of the initial load

  1. . Query 1: Select a year
  2. . Within the year, drill down to a specific country
  3. . Within the country, drill down to a specific sales organization
  4. . Within the sales organization, drill down to distribution channel
  5. . Within the distribution channel, drill down to type, version, and material number
  6. . Change material number and customer number
  7. . Start new query including formulas
  8. . Query 2: Start a new query that retrieves the yearly and quarterly sales volumes for different types of ranges, e.g. sales organizations
  9. . Query 3: Similar to query 1 for country, sales organization, and distribution channel
  10. . Within this query, expand to material level two
  11. . Expand to level three

SAP NetWeaver 2004s and SAP NetWeaver 7.0, Without BI Accelerator  300,000,000 records

Date of Certification (mm/dd/yyyy) Technology Partner Number of records Query Navigation Steps/ Hour Operating System – Release Database Server RDBMS Release SAP NetWeaver Release Number & Type of Database/Application Servers Certification Number
11/30/2010 IBM 300,000,000 625646 i 7.1 DB2 for i 7.1 SAP NetWeaver 7.0 IBM Power System 780, 4 processor / 16 cores / 64 threads, POWER7, 4100 MHz, 32 KB(D) + 32 KB(I) L1 cache and 256 KB L2 cache per core, 8 MB L3 cache per core 2010047
8/17/2010 IBM 300,000,000 151,426 i 7.1 DB2 for i 7.1 SAP NetWeaver 7.0 IBM Power System 740, 1 processor / 4 cores / 16 threads, POWER7, 3.30 GHz, 32 KB(D) + 32 KB(I) L1 cache and 256 KB L2 cache per core, 8 MB L3 cache per core, 64 GB main memory 2010035
4/13/2010 IBM 300,000,000 278462 i 7.1 DB2 for i 7.1 SAP NetWeaver 7.0 IBM Power System 750, 1 processor / 8 cores / 32 threads, POWER7, 3.30 GHz, 32 KB(D) + 32 KB(I) L1 cache and 256 KB L2 cache per core, 4 MB L3 cache per core, 128 GB main memory 2010014
2/8/2010 IBM 300,000,000 241526 i 6.1 DB2 for i 6.1 SAP NetWeaver 7.0 IBM Power System 750, 1 processor / 8 cores / 32 threads, POWER7, 3.30 GHz, 32 KB(I) + 32 KB(D) L1 cache and 256 KB L2 cache per core, 4 MB L3 cache per core, 128 GB main memory 2010005
10/20/2009 IBM 300,000,000 154447 i 6.1 DB2 for i 6.1 SAP NetWeaver 7.0 IBM Power System 570, 4 processors / 8 cores / 16 threads, POWER6, 4.2 GHz, 128 KB L1 cache and 4 MB L2 cache per core, 32 MB L3 cache per processor, 128 GB main memory 2009043
4/28/2009 IBM 300,000,000 90,635 i 6.1 DB2 for i 6.1 SAP NetWeaver 7.0 IBM Power 550, 2 processors / 4 cores / 8 threads, POWER6, 5.0 GHz, 128 KB L1 cache and 4 MB L2 cache per core, 32 MB L3 cache per processor, 64 GB main memory 2009011
4/28/2009 IBM 300,000,000 90,492 i 6.1 DB2 for i 6.1 SAP NetWeaver 7.0 IBM Power 550, 2 processors / 4 cores / 8 threads, POWER6, 5.0 GHz, 128 KB L1 cache and 4 MB L2 cache per core, 32 MB L3 cache per processor, 64 GB main memory 2009010
3/19/2009 IBM 300,000,000 41,090 i 6.1 DB2 for i 6.1 SAP NetWeaver 7.0 IBM Power 520, 1 processor / 2 cores / 4 threads, POWER6, 4.7 GHz, 128 KB L1 cache and 4 MB L2 cache per core, 32 MB L3 cache per processor, 32 GB main memory 2009002
12/8/2008 IBM 300,000,000 93.468 i 6.1 DB2 for i 6.1 SAP NetWeaver 7.0 IBM Power System 570, 2 processors / 4 cores / 8 threads, POWER6, 5.0 GHz, 128 KB L1 cache and 4 MB L2 cache per core, 32 MB L3 cache per processor, 96 GB main memory 2008069

Enlarge
View table

SAP BW edition for SAP HANA Benchmark Version 1-3

SAP BW edition for SAP HANA Benchmark SAP BW edition for SAP HANA Benchmark Version 2 SAP BW edition for SAP HANA Benchmark Version 3

The SAP BW edition for SAP HANA Standard Application Benchmark is the latest addition to the list of SAP BW benchmarks. The new SAP BW edition for SAP HANA benchmarks goes well beyond the scope and features used in the BW-AML benchmark. The new benchmark can fully utilize the new capabilities of SAP HANA to process the benchmark workload – thus the benchmark will only be available on the SAP BW edition for SAP HANA. The benchmark workload completely differs from that of the BW-AML benchmark.

The benchmark consists of 3 phases.

  1. . Data load phase
  2. . Query throughput phase
  3. . Query runtime phase

Allowed data volumes are a multiple of 1.3 billion initial records.

Initial Data volumes: n * 1.3 billion records
Single- and Multi-Node Setups

Important: Due to the completely different workload, data model, query design and KPIs the BW-AML and SAP BW edition for SAP HANA benchmark results are not comparable.

Data load phase

The first phase of the new BW edition for SAP HANA benchmark is the load phase. The initial data of n-times 1.3 billion records is loaded into the system. 

Query Phases (Query throughput phase / Query runtime phase)

  • Inserts / Updates / Deletes on transactional data during delta loads
  • Master data changes along with delta loads
  • Utilization of navigation attributes
  • Multiple currencies in source data
  • Stand-alone data loads to measure data load times (Phase 2)
  • Master data lookups, non-trivial transformations (Phase 2)

The SAP BW edition for SAP HANA Standard Application Benchmark Version 2 is the latest addition to the list of SAP BW benchmarks. The new SAP BW edition for SAP HANA benchmarks goes well beyond the scope and features used in the BW-AML benchmark. The new benchmark can fully utilize the new capabilities of SAP HANA to process the benchmark workload – thus the benchmark will only be available on the SAP BW edition for SAP HANA. The benchmark workload completely differs from that of the BW-AML benchmark.

The benchmark consists of 3 phases.

  1. . Data load phase
  2. . Query throughput phase
  3. . Query runtime phase

Allowed data volumes are a multiple of 1.3 billion initial records.

Initial Data volumes: n * 1.3 billion records
Single- and Multi-Node Setups

Important:

  • Due to the completely different workload, data model, query design and KPIs the BW-AML and SAP BW edition for SAP HANA benchmark results are not comparable.
  • Due to differenes in the data model, the SAP BW edition for SAP HANA benchmark and SAP BW edition for SAP HANA benchmark Version 2 results are not comparable.

The SAP BW edition for SAP HANA Standard Application Benchmark Version 3 is the latest addition to the list of SAP BW benchmarks. The new benchmark can fully utilize the new capabilities of SAP HANA to process the benchmark workload – thus the benchmark will only be available on the SAP BW edition for SAP HANA.

The benchmark consists of 3 phases.

  1. . Data load phase
  2. . Query throughput phase
  3. . Query runtime phase

Allowed data volumes are a multiple of 1.3 billion initial records.

Initial Data volumes: n * 1.3 billion records
Scale-Up and Scale-Out configurations

Important:

  • Due to the completely different workload, data model, query design and KPIs the BW-AML and SAP BW edition for SAP HANA Benchmark results are not comparable.
  • Due to differenes in data model and query definitions the several  versions of SAP BW edition for SAP HANA Benchmarks are not comparable
  • SAP BW edition for SAP HANA Standard Application Benchmark Version 3 is available for SAP HANA 2.0 only
  • The Data Load KPI (Phase 1) has been changed with Version 3. With Version 3 only the runtime of the latest data set (1,3 billion records) will be the KPI.

Employee Self-Service Portal (EP-ESS)

The Employee Self-Service Portal (EP-ESS) Benchmark is part of the SAP Standard Application Benchmark suite for SAP NetWeaver Portal. 

The suite consists of two representative business cases that are highly visible in the market and allow for the testing of high numbers of concurrent users on the portal. Each scenario is a standard application benchmark in its own right and can be run separately.

User Interaction Steps of the EP-ESS Benchmark

  1. . Log on to the portal – Welcome page (includes three iViews)
  2. . Choose Working Time —> Record Working Time
  3. . Choose Leave Request
  4. . Choose Leave Request Overview
  5. . Choose Personal Information —> Personal Data
  6. . Choose Address
  7. . Choose Bank Information
  8. . Choose Benefits and Payment —> Paycheck Inquiry
  9. . Log off from portal

 

Seven different business transactions are launched per loop (steps 2 - 8). Every loop is repeated n times. The user think time is 10 seconds; thus, one loop takes at least 70 seconds.

Benchmark Results

The EP-ESS Benchmark uses the Business Package for Employee Self-Service. EP-ESS provides a number of iViews that allow employees to create, display, and change their own data, such as department, name, address, and more in the SAP R/3 back-end system. This benchmark focuses on concurrent users with the employee role (workset Employee Self-Service) and their top-level navigation behavior in the portal. From a technical point of view, this benchmark tests the performance of the portal platform (running on the SAP NetWeaver Java stack), while launching business transactions. The back-end influence on the performance has been omitted.

SAP NetWeaver Portal 6.0 and 7.0

Date of Certification (mm/dd/yyyy) Technology Partner Number of Portal Users Average Online Response Time (sec) Number of Top Level Navigation Steps/Hour Operating System - Release JSE - Version SAP NetWeaver Release Central Server Number of processors Number of cores Number of threads CPU Type CPU Speed (MHz) Cache Central Server Memory (MB) Certification Number
8/30/2007 IBM 2600 1.81 792619 AIX 5L Version 5.3 IBM Java 1.4.2 7.0 (2004s) IBM System p 570 2 4 8 POWER6 4700 128 KB L1 cache and 4 MB L2 cache per core, 32 MB L3 cache per processor 32768 2007052
4/7/2006 Dell 1941 1.99 583000 SuSE Linux Enterprise Server 9 (64-bit) IBM JSE 1.4.2 (64-bit) 2004 (64-bit) Dell PowerEdge Model 6850 4 8 16 Intel XEON Dual-Core 3000 16 KB L1 cache, 2 MB L2 cache 32768 2006017
03/02/2006 Sun 1440 1.94 433912 SuSE Linux Enterprise Server 9 (64-bit) IBM JSE 1.4.2 (64-bit) 2004 (64-bit) Sun Fire Model V20z 2 4 4 Dual-core AMD Opteron processor Model 275 2200 128 KB L1 cache, 1 MB L2 cache 8192 2006012

Enlarge
View table

People-Centric CRM Portal (EP-PCC)

The People-Centric CRM Portal (EP-PCC) Benchmark is part of the SAP Standard Application Benchmark suite for SAP NetWeaver Portal.

The suite consists of two representative business cases that are highly visible in the market and allow for the testing of high numbers of concurrent users on the portal. Each scenario is a standard application benchmark in its own right and can be run separately.

The People-Centric CRM (EP-PCC) Benchmark uses the Business Package for SAP CRM: Business Productivity Pack. This business package provides a number of iViews that display sales information from SAP Business Information Warehouse (SAP BW) and SAP Customer Relationship Management (SAP CRM) systems for employees in sales, marketing, and service who need to make the correct decision quickly and effectively at all times. This benchmark focuses on concurrent users in the sales representative role and their top-level navigation behavior in the portal. From a technical point of view, this benchmark tests the performance of the portal platform (running on the SAP NetWeaver Java stack), while launching business transactions, rendering the presentation output, and caching SAP BW business content. The SAP BW and SAP CRM back-end responses are simulated to avoid the influence of back-end load on the portal performance.

User Interaction Steps of the EP-PCC Benchmark

  1. . Log on to the portal – Welcome Page (includes three iViews)
  2. . Choose Home —> Today (includes four iViews)
  3. . Choose Activities —> Overview (includes four iViews)
  4. . Choose Account Management —> Overview (includes two iViews)
  5. . Choose Acquisition —> Overview (includes three iViews)
  6. . Log off from portal

Thirteen different iViews are rendered per loop (steps 2 - 5). Every loop is repeated n times. The user think time is 10 seconds; thus, one loop takes at least 40 seconds.

Back to top