SPEChpc™ 2021 Large Result

Copyright 2021-2023 Standard Performance Evaluation Corporation

Dell Inc. (Test Sponsor: Texas Advanced Computing Center)

Frontera: PowerEdge C6420 (Intel Xeon Platinum 8280)

SPEChpc 2021_lrg_base = 8.47

SPEChpc 2021_lrg_peak = Not Run

hpc2021 License: 6340 Test Date: Sep-2021
Test Sponsor: Texas Advanced Computing Center Hardware Availability: Jun-2019
Tested by: Texas Advanced Computing Center Software Availability: Dec-2020

Benchmark result graphs are available in the PDF report.

Results Table

Benchmark Base Peak
Model Ranks Thrds/Rnk Seconds Ratio Seconds Ratio Seconds Ratio Model Ranks Thrds/Rnk Seconds Ratio Seconds Ratio Seconds Ratio
SPEChpc 2021_lrg_base 8.47
SPEChpc 2021_lrg_peak Not Run
Results appear in the order in which they were run. Bold underlined text indicates a median measurement.
805.lbm_l OMP 1024 27 275 9.92 277 9.85 276 9.89
818.tealeaf_l OMP 1024 27 221 6.55 221 6.56 219 6.63
819.clvleaf_l OMP 1024 27 305 6.89 307 6.85 307 6.85
828.pot3d_l OMP 1024 27 701 6.49 704 6.47 704 6.46
834.hpgmgfv_l OMP 1024 27 307 10.90 304 11.00 307 10.90
835.weather_l OMP 1024 27 278 12.30 291 11.80 292 11.70
Hardware Summary
Type of System: Homogenous Cluster
Compute Node: PowerEdge C6420
Interconnect: InfiniBand
Compute Nodes Used: 512
Total Chips: 1024
Total Cores: 28672
Total Threads: 28672
Total Memory: 96 TB
Software Summary
Compiler: C/C++/Fortran: Version 2020 Update 4 of
Intel Compilers for Linux
MPI Library: Intel MPI Library 2019 Update 9 for Linux
Other MPI Info: None
Other Software: None
Base Parallel Model: OMP
Base Ranks Run: 1024
Base Threads Run: 27
Peak Parallel Models: Not Run

Node Description: PowerEdge C6420

Hardware
Number of nodes: 512
Uses of the node: compute
Vendor: Dell Inc.
Model: PowerEdge C6420
CPU Name: Intel Xeon Platinum 8280
CPU(s) orderable: 1 chips
Chips enabled: 2
Cores enabled: 56
Cores per chip: 28
Threads per core: 1
CPU Characteristics: Turbo up to 4.0 GHz
CPU MHz: 2700
Primary Cache: 32 KB I + 32 KB D on chip per core
Secondary Cache: 1 MB I+D on chip per core
L3 Cache: 38.5 MB I+D on chip per chip
Other Cache: None
Memory: 192 GB (12 x 16 GB 2Rx8 PC4-2933Y-R)
Disk Subsystem: 1x 240GB SATA SSD
Other Hardware: None
Accel Model:
Accel ECC enabled:
Accel Description:
Adapter: NVIDIA ConnectX-6 VPI Infiniband Adapter Card
Number of Adapters: 1
Slot Type: PCIe 3.0 x16
Data Rate: 100Gb/s
Ports Used: 1
Interconnect Type: Infiniband HDR100
Software
Adapter: NVIDIA ConnectX-6 VPI Infiniband Adapter Card
Adapter Driver: 5.1-2.5.8.0
Adapter Firmware: 20.25.7020
Operating System: CentOS Linux release 7.8.2003
3.10.0-1127.19.1.el7.x86_64
Local File System: xfs
Shared File System: 10.6 PB Lustre (DDN SFA18K) over Infiniband HDR100
System State: Multi-user, run level 3
Other Software: None

Interconnect Description: InfiniBand

Hardware
Vendor: NVIDIA
Model: NVIDIA Infiniband HDR
Switch Model: Quantum CS8500 HDR Modular Switch
Number of Switches: 6
Number of Ports: 600
Data Rate: 200 Gb/s
Firmware: 27.2000.1386
Switch Model: Quantum QM8790 HDR Edge Switch
Number of Switches: 202
Number of Ports: 40
Data Rate: 200 Gb/s
Firmware: 27.2008.2102
Topology: Fat Tree (blocking factor 22:18)
Primary Use: MPI traffic and Lustre access
Software

General Notes

Full HDR connectivity between switches and HDR100
connectivity to the compute nodes. Half of nodes
in a rack (44) connect to      22 downlinks of a
chassis switch as pairs of HDR100 links into
HDR200 ports of the chassis switch. The other 18
ports are uplinks to the six central switches.

Submit Notes

The config file option 'submit' was used.
    mpirun -np $ranks -ppn 2 $command

General Notes

Environment settings:
     ulimit -s unlimited

Compiler Version Notes

==============================================================================
 CC  805.lbm_l(base) 818.tealeaf_l(base) 834.hpgmgfv_l(base)
------------------------------------------------------------------------------
Intel(R) C Intel(R) 64 Compiler for applications running on Intel(R) 64,
  Version 19.1.3.304 Build 20200925_000000
Copyright (C) 1985-2020 Intel Corporation.  All rights reserved.
------------------------------------------------------------------------------

==============================================================================
 FC  819.clvleaf_l(base) 828.pot3d_l(base) 835.weather_l(base)
------------------------------------------------------------------------------
Intel(R) Fortran Intel(R) 64 Compiler for applications running on Intel(R)
  64, Version 19.1.3.304 Build 20200925_000000
Copyright (C) 1985-2020 Intel Corporation.  All rights reserved.
------------------------------------------------------------------------------

Base Compiler Invocation

C benchmarks:

 mpiicc 

Fortran benchmarks:

 mpiifort 

Base Portability Flags

805.lbm_l:  -std=gnu11 
818.tealeaf_l:  -std=gnu11 
834.hpgmgfv_l:  -std=gnu11 

Base Optimization Flags

C benchmarks:

 -O3   -no-prec-div   -fp-model fast=2   -xCORE-AVX512   -ipo   -qopenmp   -ansi-alias 

Fortran benchmarks:

 -O3   -no-prec-div   -fp-model fast=2   -xCORE-AVX512   -ipo   -qopenmp 

The flags file that was used to format this result can be browsed at
http://www.spec.org/hpc2021/flags/Intel-ic2021-official-linux64_revA.2021-10-20.00.html.

You can also download the XML flags source by saving the following link:
http://www.spec.org/hpc2021/flags/Intel-ic2021-official-linux64_revA.2021-10-20.00.xml.