CONVERGE Performance with AMD EPYC™ 7002 Series Processors

This article was written by Martin Feyereisen & Joshua Weage of the Dell EMC HPC & AI Innovation Lab in October 2019.
 

---

Table of Contents

1. Introduction
2. Benchmark System Configuration
3. Software Versions
4. CONVERGE Performance
5. Conclusion

 

---

Introduction

This blog discusses the performance of CONVERGE, a popular Computational Fluid Dynamics (CFD) application from Convergent Science on the Dell EMC Ready Solutions for HPC Digital Manufacturing with AMD EPYC™ 7002 series processors. This Dell EMC Ready Solutions for HPC was designed and configured specifically for Digital Manufacturing workloads, where Computer-Aided Engineering (CAE) applications are critical for virtual product development.
The Dell EMC Ready Solutions for HPC Digital Manufacturing uses a flexible building block approach to HPC system design, where individual building blocks can be combined to build HPC systems that are optimized for customer-specific workloads and use cases.
The Dell EMC Ready Solutions for HPC Digital Manufacturing is one of many solutions in the Dell EMC HPC solutions portfolio. Please visit www.dellemc.com/hpc for a comprehensive overview of the HPC solutions offered by Dell EMC.

 

---

Benchmark System Configuration

Performance benchmarking was performed using both 7001 and 7002 series AMD EPYC processors. The system configurations used for the performance benchmarking are shown in Table 1 and Table 2. The BIOS configuration used for the benchmarking systems is shown in Table 3.
 

 

 

---

Software Versions

Application software versions are as described in Table 4.

---

CONVERGE Performance

CONVERGE is a Computational Fluid Dynamics (CFD) tool from Convergent Science commonly used across a very wide range of CFD and multi-physics applications.  CONVERGE CFD features autonomous meshing capabilities that eliminate the grid generation bottleneck from the simulation process.  CONVERGE is the industry leader in virtual combustion research and analysis.  CFD applications typically scale well across multiple processor cores and servers, have modest memory capacity requirements, and typically perform minimal disk I/O while in the solver section. Figure 1 shows the measured performance of four standards CONVERGE benchmarks contained within the software distribution Example_cases on a dual-processor single server.  The benchmarks include: Internal_Combustion_Engines/Gasoline_spark_ignition_GDI/Tumble_GDI_SAGE(GDI-SAGE), Internal_Combustion_Engines/Heavy_Duty_Diesel/ Engine_sector_Diesel_SAGE(HDD-SAGE), Gas_Turbines/LDI_Liquid_Fuel/Gas_turbine_lean_direct_inject_LES(GTLDI-ES), and Fuel_Injectors_and_Sprays/Spray_Studies/ECN_sprayH_Lagrangian_RANS(FISsH).  The performance for each benchmark is measured using the total simulation wall clock time.

The results in Figure 1 are plotted relative to the performance of a single server configured with 24-core AMD EPYC 7451 processors (24 processor CPU).  Larger values indicate better overall performance. These results show the performance advantage available with 7002 series AMD EPYC processors.  The benchmarks were carried out on five different single server systems including the 7451(24-core), 7601(32-core), 7402(24-core), 7502(32-core), and 7702(64-core) processors.  The 32-core AMD EPYC 7502 processor provides very good performance for these benchmarks. The 64-core AMD EPYC 7702 provides a noticeable advantage over the 32-core processor results.  Overall, the 7002 "Rome" series provides a significant performance gain over the 7001 "Naples" predecessors.

---

Conclusion

The results presented in this blog show that 7002 series AMD EPYC processors offer a significant performance improvement for CONVERGE relative to 7001 series AMD EPYC processors.
 

---