How To Calculate Saps In Sap System

Calculate the SAP Application Performance Standard (SAPS) for your system configuration

Comprehensive Guide: How to Calculate SAPS in SAP System

The SAP Application Performance Standard (SAPS) is a hardware-independent unit that describes the performance of a system configuration in the SAP environment. One SAPS is defined as 2,000 fully business processed order line items per hour. Understanding how to calculate SAPS is crucial for proper system sizing, hardware selection, and performance optimization in SAP landscapes.

What is SAPS and Why is it Important?

SAPS provides a standardized way to measure and compare the performance of different hardware configurations running SAP applications. The metric was introduced to:

• Create a common benchmark for hardware vendors
• Simplify system sizing for SAP implementations
• Provide performance comparisons between different hardware configurations
• Help customers make informed decisions about hardware purchases

The SAPS Calculation Formula

The basic formula for calculating SAPS is:

SAPS = (CPU Factor × Number of Cores × Core Frequency) + (RAM Factor × RAM Size) + (Storage Factor × Storage Type)

Where:

• CPU Factor: Varies by CPU architecture (0.8-1.2 for modern processors)
• RAM Factor: Typically 0.001-0.002 per GB
• Storage Factor: 5 for NVMe, 3 for SATA SSD, 1 for HDD

Key Components Affecting SAPS Calculation

1. CPU Performance

The CPU is the most critical factor in SAPS calculation, typically accounting for 70-80% of the total SAPS value. Modern multi-core processors can significantly boost SAPS ratings:

CPU Type	Base SAPS per Core	Turbo Boost Factor
Intel Xeon Platinum 8380	125	1.35
AMD EPYC 7763	132	1.40
Intel Xeon Gold 6330	118	1.30
AMD EPYC 7543	125	1.35

2. Memory Configuration

RAM affects SAPS through:

• Capacity: More RAM allows for larger working datasets (SAP recommends 4GB RAM per SAPS for most workloads)
• Speed: DDR4-3200 provides ~5% better performance than DDR4-2666 in SAP environments
• Architecture: NUMA configurations can impact performance by 10-15%

3. Storage Subsystem

Storage performance impacts SAPS through:

Storage Type	IOPS (4K Random Read)	Latency (ms)	SAPS Impact Factor
NVMe SSD (PCIe 4.0)	500,000+	<0.1	1.0
SATA SSD	90,000	0.2-0.5	0.7
15K RPM HDD	300	5-10	0.3

Step-by-Step SAPS Calculation Process

1. Determine Your Workload Requirements

   Identify whether your primary workload is:

   • OLTP: High transaction volume (e.g., order processing, financial postings)
   • OLAP: Complex analytics and reporting
   • Mixed: Combination of both

   OLTP workloads typically require 20-30% more SAPS than OLAP for the same user count.

2. Estimate Concurrent Users

   Use the following guidelines for user estimation:

   User Type	SAPS per User (OLTP)	SAPS per User (OLAP)
   Light (Data entry, inquiries)	10	8
   Medium (Complex transactions)	25	18
   Heavy (Development, admin)	50	35

3. Calculate Base SAPS Requirement

   Multiply your estimated concurrent users by the SAPS per user value:

   Base SAPS = (Light Users × 10) + (Medium Users × 25) + (Heavy Users × 50)

4. Apply Workload Adjustments

   Adjust for your specific workload mix:

   • Pure OLTP: +25%
   • Pure OLAP: -10%
   • Mixed (default): no adjustment
5. Add Growth Buffer

   SAP recommends adding:

   • 30% for 1-2 year planning horizon
   • 50% for 3-5 year planning horizon
6. Compare with Hardware SAPS Ratings

   Consult the official SAP benchmark results to find hardware that meets or exceeds your calculated SAPS requirement.

Advanced Considerations for SAPS Calculation

Virtualization Impact

Virtualized environments typically require 10-15% additional SAPS due to:

• Hypervisor overhead (2-5%)
• Resource sharing contention
• Network virtualization overhead

For VMware environments, SAP publishes specific SAPS ratings for virtual machines.

HA/DR Configurations

High availability and disaster recovery setups require additional SAPS:

Configuration	Additional SAPS Required
Active/Passive HA	10-15%
Active/Active HA	25-30%
Disaster Recovery Site	50-100% (depending on RTO/RPO)

Cloud Considerations

Cloud providers publish SAP-certified instances with SAPS ratings:

• AWS: EC2 instances with SAP certification (e.g., m5.4xlarge = 8,500 SAPS)
• : M-series and E-series VMs with published SAPS values
• Google Cloud: n2-standard and n2-highmem machines with SAP benchmarks

Always verify current SAPS ratings as cloud providers frequently update their hardware.

Common Mistakes in SAPS Calculation

1. Underestimating User Concurrency

   Many organizations calculate based on total users rather than peak concurrent users. SAP systems typically see 15-30% of total users active concurrently during peak hours.

2. Ignoring Workload Mix

   Treating all users equally without accounting for different workload intensities (light vs. heavy users) can lead to significant under-provisioning.

3. Overlooking Storage Performance

   While CPU is the primary factor, poor storage performance can bottleneck SAP systems, effectively reducing the realized SAPS by 20-40%.

4. Not Planning for Growth

   SAP implementations often grow 20-30% annually in transaction volume. Failing to account for this growth leads to premature hardware upgrades.

5. Assuming Linear Scalability

   Adding more cores doesn’t always translate to linear SAPS increases due to NUMA effects and software licensing limitations.

Tools and Resources for SAPS Calculation

The following tools can assist with SAPS calculation and hardware sizing:

• SAP Quick Sizer:

  The official SAP tool that provides hardware recommendations based on your specific SAP modules and user counts. Available at SAP Sizing.

• SAP Benchmark Results:

  Official hardware benchmarks published by SAP, showing SAPS ratings for certified hardware configurations. Updated quarterly.

• Hardware Vendor Tools:

  Most major hardware vendors (IBM, HP, Dell, Lenovo) provide SAP-specific sizing tools that incorporate their latest benchmark results.

• Cloud Provider Calculators:

  AWS, Azure, and Google Cloud offer SAP-specific calculators that map SAPS requirements to their instance types.

Real-World SAPS Calculation Examples

Example 1: Medium-Sized OLTP System

Requirements:

• 500 total users (150 concurrent)
• 70% medium users, 20% light users, 10% heavy users
• Pure OLTP workload
• 3-year planning horizon

Calculation:

1. Base SAPS = (105 × 10) + (35 × 25) + (10 × 50) = 1,075 + 875 + 500 = 2,450 SAPS
2. OLTP adjustment = 2,450 × 1.25 = 3,062 SAPS
3. Growth buffer (50%) = 3,062 × 1.5 = 4,593 SAPS

Recommended Hardware: Dual-socket server with 24 cores (Intel Xeon Platinum) rated at 5,000 SAPS

Example 2: Large OLAP System

Requirements:

• 200 total users (80 concurrent)
• 40% medium users, 50% light users, 10% heavy users
• Pure OLAP workload
• 5-year planning horizon

Calculation:

1. Base SAPS = (40 × 8) + (32 × 18) + (8 × 35) = 320 + 576 + 280 = 1,176 SAPS
2. OLAP adjustment = 1,176 × 0.9 = 1,058 SAPS
3. Growth buffer (50%) = 1,058 × 1.5 = 1,587 SAPS

Recommended Hardware: Single-socket server with 16 cores (AMD EPYC) rated at 1,800 SAPS

Optimizing Your SAP System for Maximum SAPS

To get the most performance from your hardware investment:

1. CPU Optimization
   • Enable hyper-threading for OLAP workloads (can provide 10-15% SAPS boost)
   • Disable power-saving features in BIOS for SAP systems
   • Use CPU pinning for critical SAP work processes
2. Memory Configuration
   • Use the maximum memory speed supported by your CPU
   • Configure NUMA properly for multi-socket systems
   • Allocate sufficient memory for SAP buffers (typically 30-40% of total RAM)
3. Storage Tuning
   • Use RAID 10 for transaction logs
   • Separate data and log files on different storage volumes
   • Implement proper striping for large tablespaces
4. SAP-Specific Optimizations
   • Properly size SAP work processes
   • Optimize SAP buffer parameters
   • Implement proper table partitioning for large tables

Future Trends in SAPS Calculation

The SAPS metric continues to evolve with new technologies:

• AI and Machine Learning:

  SAP is incorporating AI into its sizing tools to provide more accurate predictions based on actual usage patterns rather than just user counts.

• Containerization:

  New SAPS benchmarks are being developed for containerized SAP environments running on Kubernetes.

• Quantum Computing:

  While still in early stages, SAP is exploring how quantum computing might impact future SAPS calculations for complex analytics workloads.

• Edge Computing:

  As SAP systems extend to edge devices, new micro-SAPS metrics may emerge for IoT and edge computing scenarios.

Frequently Asked Questions About SAPS

Q: How often does SAP update its benchmark results?

A: SAP publishes new benchmark results quarterly, with major updates coinciding with new processor releases from Intel and AMD.

Q: Can I compare SAPS across different SAP versions?

A: SAPS values are generally comparable across SAP NetWeaver versions (7.0 and above), but older R/3 versions may have different characteristics.

Q: How does HANA affect SAPS calculations?

A: SAP HANA systems use a different benchmark (SAPS for HANA) that accounts for in-memory processing. HANA typically requires 30-50% fewer SAPS for the same workload compared to traditional databases.

Q: Is there a way to measure the actual SAPS of my existing system?

A: While you can’t directly measure SAPS, you can use SAP’s ST03N transaction to analyze workload and compare against benchmark systems with known SAPS ratings.

Q: How does virtualization affect SAPS requirements?

A: Virtualized environments typically require 10-15% additional SAPS to account for hypervisor overhead and resource sharing.

Conclusion

Accurately calculating SAPS is fundamental to designing a well-performing SAP system that meets your business requirements while optimizing hardware costs. By understanding the components that contribute to SAPS, following the calculation methodology, and accounting for your specific workload characteristics, you can make informed decisions about hardware selection and system architecture.

Remember that SAPS calculation is both a science and an art – while the formulas provide a solid foundation, real-world performance will depend on your specific implementation, customizations, and usage patterns. Always validate your calculations with SAP’s official tools and consider engaging SAP-certified hardware partners for complex implementations.

For the most current information, always refer to the official SAP benchmark documentation and consult with your hardware vendor’s SAP competence center.