ARMv8 Cortex-A72 testing on Ubuntu 18.04 via the Phoronix Test Suite. Tests for a future article.
Compare your own system(s) to this result file with the
Phoronix Test Suite by running the command:
phoronix-test-suite benchmark 1903298-HV-AMAZONEC210
Ubuntu 18.04 EC2 AMD EPYC, Intel Xeon, Amazon Graviton Compute
ARMv8 Cortex-A72 testing on Ubuntu 18.04 via the Phoronix Test Suite. Tests for a future article.
m5a.2xlarge:
Processor: AMD EPYC 7571 (4 Cores / 8 Threads), Motherboard: Amazon EC2 m5a.2xlarge (1.0 BIOS), Chipset: Intel 440FX 82441FX PMC, Memory: 31744MB, Disk: 54GB Amazon Elastic Block Store, Network: Amazon Elastic
OS: Ubuntu 18.04, Kernel: 4.15.0-1032-aws (x86_64), Compiler: GCC 7.3.0, File-System: ext4, System Layer: KVM
m5ad.2xlarge:
Processor: AMD EPYC 7571 (4 Cores / 8 Threads), Motherboard: Amazon EC2 m5ad.2xlarge (1.0 BIOS), Chipset: Intel 440FX 82441FX PMC, Memory: 31744MB, Disk: 300GB Amazon EC2 NVMe Instance Storage + 54GB Amazon Elastic Block Store, Network: Amazon Elastic
OS: Ubuntu 18.04, Kernel: 4.15.0-1032-aws (x86_64), Compiler: GCC 7.3.0, File-System: ext4, System Layer: KVM
m5.2xlarge:
Processor: Intel Xeon Platinum 8175M (4 Cores / 8 Threads), Motherboard: Amazon EC2 m5.2xlarge (1.0 BIOS), Chipset: Intel 440FX 82441FX PMC, Memory: 31744MB, Disk: 54GB Amazon Elastic Block Store, Network: Amazon Elastic
OS: Ubuntu 18.04, Kernel: 4.15.0-1032-aws (x86_64), Compiler: GCC 7.3.0, File-System: ext4, System Layer: KVM
m5ad.4xlarge:
Processor: AMD EPYC 7571 (8 Cores / 16 Threads), Motherboard: Amazon EC2 m5ad.4xlarge (1.0 BIOS), Chipset: Intel 440FX 82441FX PMC, Memory: 63488MB, Disk: 2 x 300GB Amazon EC2 NVMe Instance Storage + 54GB Amazon Elastic Block Store, Network: Amazon Elastic
OS: Ubuntu 18.04, Kernel: 4.15.0-1032-aws (x86_64), Compiler: GCC 7.3.0, File-System: ext4, System Layer: KVM
m5a.4xlarge:
Processor: AMD EPYC 7571 (8 Cores / 16 Threads), Motherboard: Amazon EC2 m5a.4xlarge (1.0 BIOS), Chipset: Intel 440FX 82441FX PMC, Memory: 63488MB, Disk: 54GB Amazon Elastic Block Store, Network: Amazon Elastic
OS: Ubuntu 18.04, Kernel: 4.15.0-1032-aws (x86_64), Compiler: GCC 7.3.0, File-System: ext4, System Layer: KVM
m5.4xlarge:
Processor: Intel Xeon Platinum 8175M (8 Cores / 16 Threads), Motherboard: Amazon EC2 m5.4xlarge (1.0 BIOS), Chipset: Intel 440FX 82441FX PMC, Memory: 63488MB, Disk: 54GB Amazon Elastic Block Store, Network: Amazon Elastic
OS: Ubuntu 18.04, Kernel: 4.15.0-1032-aws (x86_64), Compiler: GCC 7.3.0, File-System: ext4, System Layer: KVM
a1.2xlarge:
Processor: ARMv8 Cortex-A72 (8 Cores), Motherboard: Amazon EC2 a1.2xlarge (1.0 BIOS), Chipset: Amazon Device 0200, Memory: 16384MB, Disk: 54GB Amazon Elastic Block Store, Network: Amazon Elastic
OS: Ubuntu 18.04, Kernel: 4.15.0-1028-aws (aarch64), Compiler: GCC 7.3.0, File-System: ext4
a1.4xlarge:
Processor: ARMv8 Cortex-A72 (16 Cores), Motherboard: Amazon EC2 a1.4xlarge (1.0 BIOS), Chipset: Amazon Device 0200, Memory: 31744MB, Disk: 54GB Amazon Elastic Block Store, Network: Amazon Elastic
OS: Ubuntu 18.04, Kernel: 4.15.0-1028-aws (aarch64), Compiler: GCC 7.3.0, File-System: ext4
7-Zip Compression 16.02
Compress Speed Test
MIPS > Higher Is Better
m5a.2xlarge .. 14382 |=====================
m5ad.2xlarge . 14838 |=====================
m5.2xlarge ... 20988 |==============================
m5ad.4xlarge . 28189 |========================================
m5a.4xlarge .. 28252 |=========================================
m5.4xlarge ... 39682 |=========================================================
a1.2xlarge ... 13432 |===================
a1.4xlarge ... 25266 |====================================
7-Zip Compression 16.02
Performance / Cost - Compress Speed Test
MIPS Per Dollar > Higher Is Better
m5a.2xlarge .. 1438200.00 |====================================
m5ad.2xlarge . 741900.00 |==================
m5.2xlarge ... 2098800.00 |====================================================
m5ad.4xlarge . 704725.00 |=================
m5a.4xlarge .. 941733.33 |=======================
m5.4xlarge ... 1322733.33 |=================================
a1.2xlarge ... 1343200.00 |=================================
a1.4xlarge ... 1263300.00 |===============================
asmFish 2018-07-23
1024 Hash Memory, 26 Depth
Nodes/second > Higher Is Better
m5a.2xlarge .. 7980807 |=====================
m5ad.2xlarge . 8328787 |======================
m5.2xlarge ... 10175757 |===========================
m5ad.4xlarge . 15994554 |==========================================
m5a.4xlarge .. 15619592 |=========================================
m5.4xlarge ... 20605138 |======================================================
a1.2xlarge ... 7565800 |====================
a1.4xlarge ... 15743121 |=========================================
asmFish 2018-07-23
Performance / Cost - 1024 Hash Memory, 26 Depth
Nodes/second Per Dollar > Higher Is Better
m5a.2xlarge .. 79808070.00 |=======================
m5ad.2xlarge . 75716245.45 |======================
m5.2xlarge ... 113063966.67 |================================
m5ad.4xlarge . 84181863.16 |========================
m5a.4xlarge .. 91879952.94 |==========================
m5.4xlarge ... 158501061.54 |=============================================
a1.2xlarge ... 126096666.67 |====================================
a1.4xlarge ... 174923566.67 |==================================================
CloverLeaf
Lagrangian-Eulerian Hydrodynamics
Seconds < Lower Is Better
m5a.2xlarge .. 5.88 |==============================================
m5ad.2xlarge . 5.48 |===========================================
m5.2xlarge ... 3.27 |=========================
m5ad.4xlarge . 4.71 |=====================================
m5a.4xlarge .. 4.74 |=====================================
m5.4xlarge ... 1.87 |===============
a1.2xlarge ... 7.47 |==========================================================
a1.4xlarge ... 5.68 |============================================
CloverLeaf
Performance / Cost - Lagrangian-Eulerian Hydrodynamics
Seconds x Dollar < Lower Is Better
m5ad.4xlarge . 0.05 |==========================================================
Cost To Run Tests
Cost / Price Per Hour
Cost / Price Per Hour < Lower Is Better
m5a.2xlarge .. 0.64 |===================================
m5ad.2xlarge . 0.73 |========================================
m5.2xlarge ... 0.51 |============================
m5ad.4xlarge . 1.07 |==========================================================
m5a.4xlarge .. 0.90 |=================================================
m5.4xlarge ... 0.75 |=========================================
a1.2xlarge ... 0.40 |======================
a1.4xlarge ... 0.57 |===============================
GraphicsMagick 1.3.30
Operation: Resizing
Iterations Per Minute > Higher Is Better
m5a.2xlarge .. 142 |==============================================
m5ad.2xlarge . 161 |====================================================
m5.2xlarge ... 164 |=====================================================
m5ad.4xlarge . 178 |==========================================================
m5a.4xlarge .. 178 |==========================================================
m5.4xlarge ... 181 |===========================================================
a1.2xlarge ... 132 |===========================================
a1.4xlarge ... 146 |================================================
GraphicsMagick 1.3.30
Performance / Cost - Operation: Resizing
Iterations Per Minute Per Dollar > Higher Is Better
m5a.2xlarge .. 7100.00 |=============================
m5ad.2xlarge . 8050.00 |=================================
m5.2xlarge ... 8200.00 |==================================
m5ad.4xlarge . 4450.00 |==================
m5a.4xlarge .. 4450.00 |==================
m5.4xlarge ... 4525.00 |===================
a1.2xlarge ... 13200.00 |======================================================
a1.4xlarge ... 7300.00 |==============================
GraphicsMagick 1.3.30
Operation: Rotate
Iterations Per Minute > Higher Is Better
m5a.2xlarge .. 152 |==================================================
m5ad.2xlarge . 179 |==========================================================
m5.2xlarge ... 179 |==========================================================
m5ad.4xlarge . 179 |==========================================================
m5a.4xlarge .. 179 |==========================================================
m5.4xlarge ... 181 |===========================================================
a1.2xlarge ... 137 |=============================================
a1.4xlarge ... 137 |=============================================
GraphicsMagick 1.3.30
Performance / Cost - Operation: Rotate
Iterations Per Minute Per Dollar > Higher Is Better
m5a.2xlarge .. 7600.00 |==============================
m5ad.2xlarge . 8950.00 |===================================
m5.2xlarge ... 8950.00 |===================================
m5ad.4xlarge . 4475.00 |==================
m5a.4xlarge .. 4475.00 |==================
m5.4xlarge ... 4525.00 |==================
a1.2xlarge ... 13700.00 |======================================================
a1.4xlarge ... 6850.00 |===========================
Himeno Benchmark 3.0
Poisson Pressure Solver
MFLOPS > Higher Is Better
m5a.2xlarge .. 144 |====
m5ad.2xlarge . 194 |=====
m5.2xlarge ... 2167 |=========================================================
m5ad.4xlarge . 905 |========================
m5a.4xlarge .. 913 |========================
m5.4xlarge ... 2193 |==========================================================
a1.2xlarge ... 488 |=============
a1.4xlarge ... 499 |=============
Himeno Benchmark 3.0
Performance / Cost - Poisson Pressure Solver
MFLOPS Per Dollar > Higher Is Better
m5a.2xlarge .. 7181.50 |====
m5ad.2xlarge . 9688.50 |=====
m5.2xlarge ... 108365.00 |=====================================================
m5ad.4xlarge . 22635.00 |===========
m5a.4xlarge .. 22815.75 |===========
m5.4xlarge ... 54826.75 |===========================
a1.2xlarge ... 48763.00 |========================
a1.4xlarge ... 24964.50 |============
LeelaChessZero 0.20.1
Backend: BLAS
Nodes Per Second > Higher Is Better
m5a.2xlarge .. 45.58 |=============
m5ad.2xlarge . 49.64 |==============
m5.2xlarge ... 133.96 |=====================================
m5ad.4xlarge . 74.84 |=====================
m5a.4xlarge .. 76.63 |=====================
m5.4xlarge ... 201.64 |========================================================
a1.2xlarge ... 42.16 |============
a1.4xlarge ... 63.95 |==================
LeelaChessZero 0.20.1
Performance / Cost - Backend: BLAS
Nodes Per Second Per Dollar > Higher Is Better
m5a.2xlarge .. 379.83 |======
m5ad.2xlarge . 354.57 |======
m5.2xlarge ... 2679.20 |============================================
m5ad.4xlarge . 415.78 |=======
m5a.4xlarge .. 510.87 |========
m5.4xlarge ... 3360.67 |=======================================================
a1.2xlarge ... 527.00 |=========
a1.4xlarge ... 639.50 |==========
NAMD 2.13b1
ATPase Simulation - 327,506 Atoms
days/ns < Lower Is Better
m5a.2xlarge .. 7.04382 |=======================================================
m5ad.2xlarge . 6.95750 |======================================================
m5.2xlarge ... 5.12792 |========================================
m5ad.4xlarge . 3.49797 |===========================
m5a.4xlarge .. 3.52828 |============================
m5.4xlarge ... 2.56888 |====================
NAMD 2.13b1
Performance / Cost - ATPase Simulation - 327,506 Atoms
days/ns x Dollar < Lower Is Better
m5a.2xlarge .. 0.70 |=====================================================
m5ad.2xlarge . 0.77 |==========================================================
m5.2xlarge ... 0.41 |===============================
m5ad.4xlarge . 0.42 |================================
m5a.4xlarge .. 0.35 |==========================
m5.4xlarge ... 0.23 |=================
OpenSSL 1.1.1
RSA 4096-bit Performance
Signs Per Second > Higher Is Better
m5a.2xlarge .. 531 |================
m5ad.2xlarge . 533 |================
m5.2xlarge ... 943 |=============================
m5ad.4xlarge . 1069 |=================================
m5a.4xlarge .. 1065 |=================================
m5.4xlarge ... 1893 |==========================================================
a1.2xlarge ... 293 |=========
a1.4xlarge ... 586 |==================
OpenSSL 1.1.1
Performance / Cost - RSA 4096-bit Performance
Signs Per Second Per Dollar > Higher Is Better
m5a.2xlarge .. 53123.00 |==========================
m5ad.2xlarge . 53290.00 |===========================
m5.2xlarge ... 94343.00 |===============================================
m5ad.4xlarge . 53451.50 |===========================
m5a.4xlarge .. 106530.00 |=====================================================
m5.4xlarge ... 94651.50 |===============================================
a1.4xlarge ... 58607.00 |=============================
PHPBench 0.8.1
PHP Benchmark Suite
Score > Higher Is Better
m5a.2xlarge .. 316373 |===================================
m5ad.2xlarge . 365982 |========================================
m5.2xlarge ... 506795 |========================================================
m5ad.4xlarge . 366592 |========================================
m5a.4xlarge .. 366029 |========================================
m5.4xlarge ... 508371 |========================================================
a1.2xlarge ... 164159 |==================
a1.4xlarge ... 164999 |==================
PHPBench 0.8.1
Performance / Cost - PHP Benchmark Suite
Score Per Dollar > Higher Is Better
m5a.2xlarge .. 31637300.00 |================================
m5ad.2xlarge . 36598200.00 |=====================================
m5.2xlarge ... 50679500.00 |===================================================
m5ad.4xlarge . 18329600.00 |==================
m5a.4xlarge .. 18301450.00 |==================
m5.4xlarge ... 50837100.00 |===================================================
a1.2xlarge ... 16415900.00 |================
a1.4xlarge ... 8249950.00 |========
Primesieve 7.4
1e12 Prime Number Generation
Seconds < Lower Is Better
m5a.2xlarge .. 76.97 |=============================
m5ad.2xlarge . 76.30 |=============================
m5.2xlarge ... 62.15 |=======================
m5ad.4xlarge . 38.24 |==============
m5a.4xlarge .. 38.46 |==============
m5.4xlarge ... 31.06 |============
a1.2xlarge ... 148.91 |========================================================
a1.4xlarge ... 71.47 |===========================
Primesieve 7.4
Performance / Cost - 1e12 Prime Number Generation
Seconds x Dollar < Lower Is Better
m5a.2xlarge .. 1.54 |====================
m5ad.2xlarge . 2.29 |==============================
m5.2xlarge ... 1.24 |================
m5ad.4xlarge . 1.15 |===============
m5a.4xlarge .. 0.77 |==========
m5.4xlarge ... 0.62 |========
a1.2xlarge ... 4.47 |==========================================================
a1.4xlarge ... 2.14 |============================
PyBench 2018-02-16
Total For Average Test Times
Milliseconds < Lower Is Better
m5a.2xlarge .. 1960 |===========================
m5ad.2xlarge . 1650 |======================
m5.2xlarge ... 1357 |==================
m5ad.4xlarge . 1663 |=======================
m5a.4xlarge .. 1655 |=======================
m5.4xlarge ... 1349 |==================
a1.2xlarge ... 4262 |==========================================================
a1.4xlarge ... 4254 |==========================================================
PyBench 2018-02-16
Performance / Cost - Total For Average Test Times
Milliseconds x Dollar < Lower Is Better
m5a.2xlarge .. 19.60 |=========
m5ad.2xlarge . 16.50 |=======
m5.2xlarge ... 13.57 |======
m5ad.4xlarge . 49.89 |======================
m5a.4xlarge .. 33.10 |===============
m5.4xlarge ... 26.98 |============
a1.2xlarge ... 85.24 |=====================================
a1.4xlarge ... 127.62 |========================================================
Scikit-Learn 0.17.1
Seconds < Lower Is Better
m5a.2xlarge .. 24.37 |=====================================
m5ad.2xlarge . 20.76 |===============================
m5.2xlarge ... 18.16 |===========================
m5ad.4xlarge . 18.84 |============================
m5a.4xlarge .. 18.93 |=============================
m5.4xlarge ... 17.06 |==========================
a1.2xlarge ... 37.81 |=========================================================
a1.4xlarge ... 32.70 |=================================================
Scikit-Learn 0.17.1
Performance / Cost -
Seconds x Dollar < Lower Is Better
m5a.2xlarge .. 0.24 |=====================================
m5ad.2xlarge . 0.21 |================================
m5.2xlarge ... 0.18 |===========================
m5ad.4xlarge . 0.38 |==========================================================
m5a.4xlarge .. 0.19 |=============================
m5.4xlarge ... 0.17 |==========================
a1.2xlarge ... 0.38 |==========================================================
a1.4xlarge ... 0.33 |==================================================
Stockfish 9
Total Time
Nodes Per Second > Higher Is Better
m5a.2xlarge .. 7553067 |======================
m5ad.2xlarge . 7665654 |======================
m5.2xlarge ... 9311255 |===========================
m5ad.4xlarge . 15080257 |============================================
m5a.4xlarge .. 15294841 |=============================================
m5.4xlarge ... 18496508 |======================================================
Stockfish 9
Performance / Cost - Total Time
Nodes Per Second Per Dollar > Higher Is Better
m5a.2xlarge .. 151061340.00 |=================================
m5ad.2xlarge . 127760900.00 |============================
m5.2xlarge ... 186225100.00 |========================================
m5ad.4xlarge . 150802570.00 |=================================
m5a.4xlarge .. 191185512.50 |=========================================
m5.4xlarge ... 231206350.00 |==================================================
Timed HMMer Search 2.3.2
Pfam Database Search
Seconds < Lower Is Better
m5a.2xlarge .. 12.92 |=========================================================
m5ad.2xlarge . 12.69 |========================================================
m5.2xlarge ... 9.08 |========================================
m5ad.4xlarge . 9.14 |========================================
m5a.4xlarge .. 9.20 |=========================================
m5.4xlarge ... 6.08 |===========================
Timed HMMer Search 2.3.2
Performance / Cost - Pfam Database Search
Seconds x Dollar < Lower Is Better
m5ad.2xlarge . 0.13 |==========================================================
m5ad.4xlarge . 0.09 |========================================
m5a.4xlarge .. 0.09 |========================================
m5.4xlarge ... 0.06 |===========================
Timed Linux Kernel Compilation 4.18
Time To Compile
Seconds < Lower Is Better
m5a.2xlarge .. 246.00 |=====================
m5ad.2xlarge . 240.00 |====================
m5.2xlarge ... 170.00 |==============
m5ad.4xlarge . 127.00 |===========
m5a.4xlarge .. 127.00 |===========
m5.4xlarge ... 89.47 |========
a1.2xlarge ... 656.60 |========================================================
a1.4xlarge ... 378.74 |================================
Timed Linux Kernel Compilation 4.18
Performance / Cost - Time To Compile
Seconds x Dollar < Lower Is Better
m5a.2xlarge .. 17.20 |==============
m5ad.2xlarge . 21.63 |=================
m5.2xlarge ... 10.19 |========
m5ad.4xlarge . 12.70 |==========
m5a.4xlarge .. 10.13 |========
m5.4xlarge ... 5.37 |====
a1.2xlarge ... 72.23 |=========================================================
a1.4xlarge ... 49.24 |=======================================
Timed MAFFT Alignment 7.392
Multiple Sequence Alignment
Seconds < Lower Is Better
m5a.2xlarge .. 7.05 |==========================================================
m5ad.2xlarge . 6.94 |=========================================================
m5.2xlarge ... 4.65 |======================================
m5ad.4xlarge . 4.67 |======================================
m5a.4xlarge .. 4.84 |========================================
m5.4xlarge ... 3.67 |==============================
a1.2xlarge ... 6.95 |=========================================================
a1.4xlarge ... 6.10 |==================================================
Timed MAFFT Alignment 7.392
Performance / Cost - Multiple Sequence Alignment
Seconds x Dollar < Lower Is Better
m5ad.4xlarge . 0.05 |==========================================================
TTSIOD 3D Renderer 2.3b
Phong Rendering With Soft-Shadow Mapping
FPS > Higher Is Better
m5a.2xlarge .. 163 |=================================
m5ad.2xlarge . 165 |=================================
m5.2xlarge ... 153 |===============================
m5ad.4xlarge . 251 |===================================================
m5a.4xlarge .. 250 |===================================================
m5.4xlarge ... 291 |===========================================================
a1.2xlarge ... 156 |================================
a1.4xlarge ... 289 |===========================================================
TTSIOD 3D Renderer 2.3b
Performance / Cost - Phong Rendering With Soft-Shadow Mapping
FPS Per Dollar > Higher Is Better
m5a.2xlarge .. 8125.50 |============================
m5ad.2xlarge . 5503.00 |===================
m5.2xlarge ... 5094.33 |==================
m5ad.4xlarge . 6272.00 |======================
m5a.4xlarge .. 8319.00 |=============================
m5.4xlarge ... 9708.00 |==================================
a1.2xlarge ... 15571.00 |======================================================
a1.4xlarge ... 14431.50 |==================================================
Zstd Compression 1.3.4
Compressing ubuntu-16.04.3-server-i386.img, Compression Level 19
Seconds < Lower Is Better
m5a.2xlarge .. 58.05 |==================================================
m5ad.2xlarge . 54.58 |===============================================
m5.2xlarge ... 35.66 |===============================
m5ad.4xlarge . 33.49 |=============================
m5a.4xlarge .. 33.60 |=============================
m5.4xlarge ... 19.77 |=================
a1.2xlarge ... 65.99 |=========================================================
a1.4xlarge ... 40.08 |===================================
Zstd Compression 1.3.4
Performance / Cost - Compressing ubuntu-16.04.3-server-i386.img, Compression Level 19
Seconds x Dollar < Lower Is Better
m5a.2xlarge .. 1.16 |==========================================================
m5ad.2xlarge . 1.09 |=======================================================
m5.2xlarge ... 0.36 |==================
m5ad.4xlarge . 1.00 |==================================================
m5a.4xlarge .. 0.67 |==================================
m5.4xlarge ... 0.40 |====================
a1.2xlarge ... 0.66 |=================================
a1.4xlarge ... 0.40 |====================
Geometric Mean Of All Test Results
Result Composite - Ubuntu 18.04 EC2 AMD EPYC, Intel Xeon, Amazon Graviton Compute
Geometric Mean > Higher Is Better
m5a.2xlarge .. 90.56 |=======================
m5ad.2xlarge . 98.97 |=========================
m5.2xlarge ... 153.43 |======================================
m5ad.4xlarge . 152.48 |======================================
m5a.4xlarge .. 152.01 |======================================
m5.4xlarge ... 223.31 |========================================================
a1.2xlarge ... 72.03 |==================
a1.4xlarge ... 101.93 |==========================