Amazon AWS Graviton3E vs. Graviton 2/3 benchmarks Benchmarks by Michael Larabel for a future article on Phoronix.com. m7g.16xlarge Graviton3: Processor: ARMv8 Neoverse-V1 (64 Cores), Motherboard: Amazon EC2 m7g.16xlarge (1.0 BIOS), Chipset: Amazon Device 0200, Memory: 256GB, Disk: 215GB Amazon Elastic Block Store, Network: Amazon Elastic OS: Ubuntu 22.04, Kernel: 5.19.0-1025-aws (aarch64), Compiler: GCC 11.3.0, File-System: ext4, System Layer: amazon c6g.16xlarge Graviton2: Processor: ARMv8 Neoverse-N1 (64 Cores), Motherboard: Amazon EC2 c6g.16xlarge (1.0 BIOS), Chipset: Amazon Device 0200, Memory: 128GB, Disk: 215GB Amazon Elastic Block Store, Network: Amazon Elastic OS: Ubuntu 22.04, Kernel: 5.19.0-1025-aws (aarch64), Compiler: GCC 11.3.0, File-System: ext4, System Layer: amazon c7g.16xlarge Graviton3: Processor: ARMv8 Neoverse-V1 (64 Cores), Motherboard: Amazon EC2 c7g.16xlarge (1.0 BIOS), Chipset: Amazon Device 0200, Memory: 128GB, Disk: 215GB Amazon Elastic Block Store, Network: Amazon Elastic OS: Ubuntu 22.04, Kernel: 5.19.0-1025-aws (aarch64), Compiler: GCC 11.3.0, File-System: ext4, System Layer: amazon c7gn.16xlarge Graviton3E: Processor: ARMv8 Neoverse-V1 (64 Cores), Motherboard: Amazon EC2 c7gn.16xlarge (1.0 BIOS), Chipset: Amazon Device 0200, Memory: 128GB, Disk: 215GB Amazon Elastic Block Store, Network: Amazon Elastic OS: Ubuntu 22.04, Kernel: 5.19.0-1025-aws (aarch64), Compiler: GCC 11.3.0, File-System: ext4, System Layer: amazon c6a.16xlarge AMD Zen 3: Processor: AMD EPYC 7R13 (32 Cores / 64 Threads), Motherboard: Amazon EC2 c6a.16xlarge (1.0 BIOS), Chipset: Intel 440FX 82441FX PMC, Memory: 128GB, Disk: 322GB Amazon Elastic Block Store, Network: Amazon Elastic OS: Ubuntu 22.04, Kernel: 5.19.0-1025-aws (x86_64), Vulkan: 1.3.238, Compiler: GCC 11.4.0, File-System: ext4, System Layer: amazon egeo-07: Processor: 2 x Intel Xeon Silver 4208 @ 3.20GHz (16 Cores / 32 Threads), Motherboard: Dell Precision 7920 Rack 0DY2X0 (2.21.2 BIOS), Chipset: Intel Sky Lake-E DMI3 Registers, Memory: 64GB, Disk: 2000GB TOSHIBA DT01ACA2, Graphics: Matrox G200eW3 15GB, Audio: NVIDIA TU104 HD Audio, Monitor: DELL 17FP, Network: 4 x Intel I350 OS: Debian 11, Kernel: 5.10.0-28-amd64 (x86_64), Display Server: X Server, Display Driver: NVIDIA, OpenCL: OpenCL 3.0 CUDA 12.2.138, Vulkan: 1.3.242, Compiler: GCC 10.2.1 20210110 + Clang 11.0.1-2 + CUDA 11.2, File-System: ext4, Screen Resolution: 1280x1024 Pennant 1.0.1 Test: sedovbig Hydro Cycle Time - Seconds < Lower Is Better m7g.16xlarge Graviton3 ... 9.206490 |==== c6g.16xlarge Graviton2 ... 16.480500 |======== c7g.16xlarge Graviton3 ... 9.422270 |==== c7gn.16xlarge Graviton3E . 9.340953 |==== c6a.16xlarge AMD Zen 3 ... 16.530500 |======== egeo-07 .................. 89.004900 |========================================= OpenSSL 3.1 Algorithm: SHA512 byte/s > Higher Is Better m7g.16xlarge Graviton3 ... 32125448870 |======================================= c6g.16xlarge Graviton2 ... 14393925490 |================= c7g.16xlarge Graviton3 ... 32145914147 |======================================= c7gn.16xlarge Graviton3E . 32126059040 |======================================= c6a.16xlarge AMD Zen 3 ... 15291283297 |=================== egeo-07 .................. 3878566000 |===== OpenSSL 3.1 Algorithm: AES-256-GCM byte/s > Higher Is Better m7g.16xlarge Graviton3 ... 283333113630 |=============================== c6g.16xlarge Graviton2 ... 129199593157 |============== c7g.16xlarge Graviton3 ... 283373795737 |=============================== c7gn.16xlarge Graviton3E . 351152465420 |====================================== c6a.16xlarge AMD Zen 3 ... 138457889450 |=============== egeo-07 .................. 44619128560 |===== BRL-CAD 7.34 VGR Performance Metric VGR Performance Metric > Higher Is Better m7g.16xlarge Graviton3 ... 783777 |============================================ c6g.16xlarge Graviton2 ... 533020 |============================== c7g.16xlarge Graviton3 ... 789066 |============================================ c7gn.16xlarge Graviton3E . 744743 |========================================== c6a.16xlarge AMD Zen 3 ... 485038 |=========================== egeo-07 .................. 102684 |====== Stress-NG 0.15.10 Test: Vector Shuffle Bogo Ops/s > Higher Is Better m7g.16xlarge Graviton3 ... 54143.40 |========================================== c6g.16xlarge Graviton2 ... 35614.51 |=========================== c7g.16xlarge Graviton3 ... 54472.07 |========================================== c7gn.16xlarge Graviton3E . 54695.04 |========================================== c6a.16xlarge AMD Zen 3 ... 22255.84 |================= egeo-07 .................. 7162.56 |====== HeFFTe - Highly Efficient FFT for Exascale 2.3 Test: c2c - Backend: FFTW - Precision: float - X Y Z: 128 GFLOP/s > Higher Is Better m7g.16xlarge Graviton3 ... 186.36 |============================================ c6g.16xlarge Graviton2 ... 135.36 |================================ c7g.16xlarge Graviton3 ... 184.03 |=========================================== c7gn.16xlarge Graviton3E . 184.11 |=========================================== c6a.16xlarge AMD Zen 3 ... 98.70 |======================= egeo-07 .................. 26.74 |====== Xcompact3d Incompact3d 2021-03-11 Input: input.i3d 129 Cells Per Direction Seconds < Lower Is Better m7g.16xlarge Graviton3 ... 3.09871038 |====== c6g.16xlarge Graviton2 ... 5.63720735 |========== c7g.16xlarge Graviton3 ... 3.14447999 |====== c7gn.16xlarge Graviton3E . 3.11489828 |====== c6a.16xlarge AMD Zen 3 ... 7.01975288 |============= egeo-07 .................. 21.23927370 |======================================= OpenSSL 3.1 Algorithm: AES-128-GCM byte/s > Higher Is Better m7g.16xlarge Graviton3 ... 332033171900 |=============================== c6g.16xlarge Graviton2 ... 158436163857 |=============== c7g.16xlarge Graviton3 ... 332064349843 |=============================== c7gn.16xlarge Graviton3E . 411130469943 |====================================== c6a.16xlarge AMD Zen 3 ... 151449269317 |============== egeo-07 .................. 61131320253 |====== Stress-NG 0.15.10 Test: Matrix Math Bogo Ops/s > Higher Is Better m7g.16xlarge Graviton3 ... 368750.67 |========================================= c6g.16xlarge Graviton2 ... 284713.63 |================================ c7g.16xlarge Graviton3 ... 368671.39 |========================================= c7gn.16xlarge Graviton3E . 369258.89 |========================================= c6a.16xlarge AMD Zen 3 ... 147576.41 |================ egeo-07 .................. 55962.00 |====== Pennant 1.0.1 Test: leblancbig Hydro Cycle Time - Seconds < Lower Is Better m7g.16xlarge Graviton3 ... 6.720537 |====== c6g.16xlarge Graviton2 ... 12.176830 |=========== c7g.16xlarge Graviton3 ... 6.961345 |======= c7gn.16xlarge Graviton3E . 6.839998 |====== c6a.16xlarge AMD Zen 3 ... 9.917565 |========= egeo-07 .................. 43.458800 |========================================= Stress-NG 0.15.10 Test: Memory Copying Bogo Ops/s > Higher Is Better m7g.16xlarge Graviton3 ... 20484.24 |========================================== c6g.16xlarge Graviton2 ... 11324.79 |======================= c7g.16xlarge Graviton3 ... 20478.67 |========================================== c7gn.16xlarge Graviton3E . 20475.96 |========================================== c6a.16xlarge AMD Zen 3 ... 8080.43 |================= egeo-07 .................. 3209.83 |======= Stress-NG 0.15.10 Test: Fused Multiply-Add Bogo Ops/s > Higher Is Better m7g.16xlarge Graviton3 ... 63762252.76 |======================================= c6g.16xlarge Graviton2 ... 37732190.54 |======================= c7g.16xlarge Graviton3 ... 63818458.61 |======================================= c7gn.16xlarge Graviton3E . 63723431.55 |======================================= c6a.16xlarge AMD Zen 3 ... 30920910.92 |=================== egeo-07 .................. 10084119.21 |====== Xcompact3d Incompact3d 2021-03-11 Input: input.i3d 193 Cells Per Direction Seconds < Lower Is Better m7g.16xlarge Graviton3 ... 13.95 |======= c6g.16xlarge Graviton2 ... 25.88 |============== c7g.16xlarge Graviton3 ... 13.83 |======= c7gn.16xlarge Graviton3E . 13.76 |======= c6a.16xlarge AMD Zen 3 ... 30.31 |================ egeo-07 .................. 84.66 |============================================= HeFFTe - Highly Efficient FFT for Exascale 2.3 Test: r2c - Backend: FFTW - Precision: double - X Y Z: 128 GFLOP/s > Higher Is Better m7g.16xlarge Graviton3 ... 138.01 |============================================ c6g.16xlarge Graviton2 ... 81.45 |========================== c7g.16xlarge Graviton3 ... 133.51 |=========================================== c7gn.16xlarge Graviton3E . 133.42 |=========================================== c6a.16xlarge AMD Zen 3 ... 86.37 |============================ egeo-07 .................. 23.02 |======= Laghos 3.1 Test: Sedov Blast Wave, ube_922_hex.mesh Major Kernels Total Rate > Higher Is Better m7g.16xlarge Graviton3 ... 410.55 |=========================================== c6g.16xlarge Graviton2 ... 322.37 |================================== c7g.16xlarge Graviton3 ... 408.01 |========================================== c7gn.16xlarge Graviton3E . 423.11 |============================================ c6a.16xlarge AMD Zen 3 ... 275.92 |============================= egeo-07 .................. 70.77 |======= HeFFTe - Highly Efficient FFT for Exascale 2.3 Test: c2c - Backend: FFTW - Precision: double - X Y Z: 128 GFLOP/s > Higher Is Better m7g.16xlarge Graviton3 ... 57.15030 |========================================== c6g.16xlarge Graviton2 ... 32.74680 |======================== c7g.16xlarge Graviton3 ... 55.10550 |======================================== c7gn.16xlarge Graviton3E . 55.10380 |======================================== c6a.16xlarge AMD Zen 3 ... 48.94320 |==================================== egeo-07 .................. 9.62296 |======= Stress-NG 0.15.10 Test: Matrix 3D Math Bogo Ops/s > Higher Is Better m7g.16xlarge Graviton3 ... 10403.93 |======================================== c6g.16xlarge Graviton2 ... 5752.17 |====================== c7g.16xlarge Graviton3 ... 10813.59 |========================================== c7gn.16xlarge Graviton3E . 10882.02 |========================================== c6a.16xlarge AMD Zen 3 ... 4571.96 |================== egeo-07 .................. 1841.58 |======= Graph500 3.0 Scale: 26 bfs max_TEPS > Higher Is Better m7g.16xlarge Graviton3 ... 1227790000 |======================================== c6g.16xlarge Graviton2 ... 874389000 |============================ c7g.16xlarge Graviton3 ... 1206990000 |======================================= c7gn.16xlarge Graviton3E . 1207760000 |======================================= c6a.16xlarge AMD Zen 3 ... 417777000 |============== egeo-07 .................. 210572000 |======= Stress-NG 0.15.10 Test: Vector Math Bogo Ops/s > Higher Is Better m7g.16xlarge Graviton3 ... 217235.59 |======================================== c6g.16xlarge Graviton2 ... 147886.14 |=========================== c7g.16xlarge Graviton3 ... 217446.12 |======================================== c7gn.16xlarge Graviton3E . 217567.10 |======================================== c6a.16xlarge AMD Zen 3 ... 221776.15 |========================================= egeo-07 .................. 38515.89 |======= NWChem 7.0.2 Input: C240 Buckyball Seconds < Lower Is Better m7g.16xlarge Graviton3 ... 1940.2 |======== c6g.16xlarge Graviton2 ... 2976.9 |============ c7g.16xlarge Graviton3 ... 1962.7 |======== c7gn.16xlarge Graviton3E . 1914.0 |======= c6a.16xlarge AMD Zen 3 ... 3440.4 |============= egeo-07 .................. 10984.0 |=========================================== srsRAN Project 23.5 Test: PUSCH Processor Benchmark, Throughput Total Mbps > Higher Is Better m7g.16xlarge Graviton3 ... 5413.8 |===================================== c6g.16xlarge Graviton2 ... 3938.7 |=========================== c7g.16xlarge Graviton3 ... 5356.8 |==================================== c7gn.16xlarge Graviton3E . 5431.2 |===================================== c6a.16xlarge AMD Zen 3 ... 6479.1 |============================================ egeo-07 .................. 1129.5 |======== Graph500 3.0 Scale: 26 bfs median_TEPS > Higher Is Better m7g.16xlarge Graviton3 ... 1194320000 |======================================== c6g.16xlarge Graviton2 ... 860432000 |============================= c7g.16xlarge Graviton3 ... 1177710000 |======================================= c7gn.16xlarge Graviton3E . 1175640000 |======================================= c6a.16xlarge AMD Zen 3 ... 410571000 |============== egeo-07 .................. 208288000 |======= HeFFTe - Highly Efficient FFT for Exascale 2.3 Test: r2c - Backend: FFTW - Precision: float - X Y Z: 128 GFLOP/s > Higher Is Better m7g.16xlarge Graviton3 ... 306.54 |============================================ c6g.16xlarge Graviton2 ... 209.50 |============================== c7g.16xlarge Graviton3 ... 301.42 |=========================================== c7gn.16xlarge Graviton3E . 300.40 |=========================================== c6a.16xlarge AMD Zen 3 ... 158.86 |======================= egeo-07 .................. 58.20 |======== LAMMPS Molecular Dynamics Simulator 23Jun2022 Model: Rhodopsin Protein ns/day > Higher Is Better m7g.16xlarge Graviton3 ... 37.558 |============================================ c6g.16xlarge Graviton2 ... 25.950 |============================== c7g.16xlarge Graviton3 ... 37.412 |============================================ c7gn.16xlarge Graviton3E . 37.482 |============================================ c6a.16xlarge AMD Zen 3 ... 19.563 |======================= egeo-07 .................. 7.176 |======== HeFFTe - Highly Efficient FFT for Exascale 2.3 Test: r2c - Backend: FFTW - Precision: float - X Y Z: 256 GFLOP/s > Higher Is Better m7g.16xlarge Graviton3 ... 164.87 |============================================ c6g.16xlarge Graviton2 ... 92.40 |========================= c7g.16xlarge Graviton3 ... 162.01 |=========================================== c7gn.16xlarge Graviton3E . 162.36 |=========================================== c6a.16xlarge AMD Zen 3 ... 102.65 |=========================== egeo-07 .................. 32.43 |========= NAS Parallel Benchmarks 3.4 Test / Class: LU.C Total Mop/s > Higher Is Better m7g.16xlarge Graviton3 ... 28341.68 |============= c6g.16xlarge Graviton2 ... 18741.90 |======== c7g.16xlarge Graviton3 ... 28375.71 |============= c7gn.16xlarge Graviton3E . 28369.11 |============= c6a.16xlarge AMD Zen 3 ... 95221.40 |========================================== egeo-07 .................. 36556.34 |================ LULESH 2.0.3 z/s > Higher Is Better m7g.16xlarge Graviton3 ... 28296.38 |========================================= c6g.16xlarge Graviton2 ... 17557.49 |========================== c7g.16xlarge Graviton3 ... 28708.66 |========================================== c7gn.16xlarge Graviton3E . 28736.23 |========================================== c6a.16xlarge AMD Zen 3 ... 16708.26 |======================== egeo-07 .................. 5676.78 |======== Stress-NG 0.15.10 Test: NUMA Bogo Ops/s > Higher Is Better m7g.16xlarge Graviton3 ... 3759.10 |=========================================== c6g.16xlarge Graviton2 ... 2112.66 |======================== c7g.16xlarge Graviton3 ... 3523.58 |======================================== c7gn.16xlarge Graviton3E . 3525.17 |======================================== c6a.16xlarge AMD Zen 3 ... 552.68 |====== egeo-07 .................. 0.89 | Remhos 1.0 Test: Sample Remap Example Seconds < Lower Is Better m7g.16xlarge Graviton3 ... 14.04 |========= c6g.16xlarge Graviton2 ... 20.74 |============== c7g.16xlarge Graviton3 ... 14.12 |========= c7gn.16xlarge Graviton3E . 14.08 |========= c6a.16xlarge AMD Zen 3 ... 22.10 |============== egeo-07 .................. 68.96 |============================================= LAMMPS Molecular Dynamics Simulator 23Jun2022 Model: 20k Atoms ns/day > Higher Is Better m7g.16xlarge Graviton3 ... 36.927 |============================================ c6g.16xlarge Graviton2 ... 25.171 |============================== c7g.16xlarge Graviton3 ... 36.862 |============================================ c7gn.16xlarge Graviton3E . 36.838 |============================================ c6a.16xlarge AMD Zen 3 ... 20.342 |======================== egeo-07 .................. 7.539 |========= Graph500 3.0 Scale: 26 sssp max_TEPS > Higher Is Better m7g.16xlarge Graviton3 ... 419754000 |========================================= c6g.16xlarge Graviton2 ... 284689000 |============================ c7g.16xlarge Graviton3 ... 415758000 |========================================= c7gn.16xlarge Graviton3E . 411762000 |======================================== c6a.16xlarge AMD Zen 3 ... 204550000 |==================== egeo-07 .................. 85755800 |======== Rodinia 3.1 Test: OpenMP LavaMD Seconds < Lower Is Better m7g.16xlarge Graviton3 ... 43.79 |========= c6g.16xlarge Graviton2 ... 62.22 |============= c7g.16xlarge Graviton3 ... 43.96 |========= c7gn.16xlarge Graviton3E . 44.04 |========= c6a.16xlarge AMD Zen 3 ... 64.18 |============= egeo-07 .................. 212.55 |============================================ 7-Zip Compression 22.01 Test: Decompression Rating MIPS > Higher Is Better m7g.16xlarge Graviton3 ... 285540 |============================================ c6g.16xlarge Graviton2 ... 234202 |==================================== c7g.16xlarge Graviton3 ... 285633 |============================================ c7gn.16xlarge Graviton3E . 285677 |============================================ c6a.16xlarge AMD Zen 3 ... 235787 |==================================== egeo-07 .................. 59055 |========= HeFFTe - Highly Efficient FFT for Exascale 2.3 Test: c2c - Backend: FFTW - Precision: float - X Y Z: 256 GFLOP/s > Higher Is Better m7g.16xlarge Graviton3 ... 81.44 |============================================= c6g.16xlarge Graviton2 ... 41.98 |======================= c7g.16xlarge Graviton3 ... 81.01 |============================================= c7gn.16xlarge Graviton3E . 81.17 |============================================= c6a.16xlarge AMD Zen 3 ... 43.59 |======================== egeo-07 .................. 17.00 |========= HeFFTe - Highly Efficient FFT for Exascale 2.3 Test: r2c - Backend: FFTW - Precision: float - X Y Z: 512 GFLOP/s > Higher Is Better m7g.16xlarge Graviton3 ... 162.96 |============================================ c6g.16xlarge Graviton2 ... 81.94 |====================== c7g.16xlarge Graviton3 ... 163.28 |============================================ c7gn.16xlarge Graviton3E . 163.56 |============================================ c6a.16xlarge AMD Zen 3 ... 82.76 |====================== egeo-07 .................. 34.37 |========= HeFFTe - Highly Efficient FFT for Exascale 2.3 Test: r2c - Backend: FFTW - Precision: double - X Y Z: 256 GFLOP/s > Higher Is Better m7g.16xlarge Graviton3 ... 78.50 |============================================= c6g.16xlarge Graviton2 ... 40.11 |======================= c7g.16xlarge Graviton3 ... 77.77 |============================================= c7gn.16xlarge Graviton3E . 78.17 |============================================= c6a.16xlarge AMD Zen 3 ... 41.59 |======================== egeo-07 .................. 17.12 |========== GPAW 23.6 Input: Carbon Nanotube Seconds < Lower Is Better m7g.16xlarge Graviton3 ... 61.83 |=========== c6g.16xlarge Graviton2 ... 92.76 |================ c7g.16xlarge Graviton3 ... 62.08 |=========== c7gn.16xlarge Graviton3E . 56.44 |========== c6a.16xlarge AMD Zen 3 ... 89.82 |=============== egeo-07 .................. 257.46 |============================================ Stress-NG 0.15.10 Test: Vector Floating Point Bogo Ops/s > Higher Is Better m7g.16xlarge Graviton3 ... 76102.55 |================================= c6g.16xlarge Graviton2 ... 42850.82 |=================== c7g.16xlarge Graviton3 ... 76178.46 |================================= c7gn.16xlarge Graviton3E . 76911.74 |================================= c6a.16xlarge AMD Zen 3 ... 96529.51 |========================================== egeo-07 .................. 21347.32 |========= HeFFTe - Highly Efficient FFT for Exascale 2.3 Test: c2c - Backend: FFTW - Precision: float - X Y Z: 512 GFLOP/s > Higher Is Better m7g.16xlarge Graviton3 ... 88.05 |============================================= c6g.16xlarge Graviton2 ... 42.83 |====================== c7g.16xlarge Graviton3 ... 88.18 |============================================= c7gn.16xlarge Graviton3E . 88.46 |============================================= c6a.16xlarge AMD Zen 3 ... 44.32 |======================= egeo-07 .................. 19.67 |========== HeFFTe - Highly Efficient FFT for Exascale 2.3 Test: r2c - Backend: FFTW - Precision: double - X Y Z: 512 GFLOP/s > Higher Is Better m7g.16xlarge Graviton3 ... 84.47 |============================================= c6g.16xlarge Graviton2 ... 44.93 |======================== c7g.16xlarge Graviton3 ... 84.75 |============================================= c7gn.16xlarge Graviton3E . 85.01 |============================================= c6a.16xlarge AMD Zen 3 ... 42.44 |====================== egeo-07 .................. 18.99 |========== Coremark 1.0 CoreMark Size 666 - Iterations Per Second Iterations/Sec > Higher Is Better m7g.16xlarge Graviton3 ... 1601880.34 |======================================== c6g.16xlarge Graviton2 ... 1260642.18 |=============================== c7g.16xlarge Graviton3 ... 1605948.67 |======================================== c7gn.16xlarge Graviton3E . 1611801.56 |======================================== c6a.16xlarge AMD Zen 3 ... 1466587.04 |==================================== egeo-07 .................. 362563.29 |========= HeFFTe - Highly Efficient FFT for Exascale 2.3 Test: c2c - Backend: FFTW - Precision: double - X Y Z: 256 GFLOP/s > Higher Is Better m7g.16xlarge Graviton3 ... 40.89230 |========================================== c6g.16xlarge Graviton2 ... 20.62790 |===================== c7g.16xlarge Graviton3 ... 40.82830 |========================================== c7gn.16xlarge Graviton3E . 40.97080 |========================================== c6a.16xlarge AMD Zen 3 ... 20.87190 |===================== egeo-07 .................. 9.32379 |========== HeFFTe - Highly Efficient FFT for Exascale 2.3 Test: c2c - Backend: FFTW - Precision: double - X Y Z: 512 GFLOP/s > Higher Is Better m7g.16xlarge Graviton3 ... 46.25 |============================================= c6g.16xlarge Graviton2 ... 24.27 |======================= c7g.16xlarge Graviton3 ... 46.37 |============================================= c7gn.16xlarge Graviton3E . 46.53 |============================================= c6a.16xlarge AMD Zen 3 ... 23.52 |======================= egeo-07 .................. 10.67 |========== Graph500 3.0 Scale: 26 sssp median_TEPS > Higher Is Better m7g.16xlarge Graviton3 ... 299497000 |========================================= c6g.16xlarge Graviton2 ... 209350000 |============================= c7g.16xlarge Graviton3 ... 293826000 |======================================== c7gn.16xlarge Graviton3E . 296164000 |========================================= c6a.16xlarge AMD Zen 3 ... 157688000 |====================== egeo-07 .................. 69070600 |========= GROMACS 2023 Implementation: MPI CPU - Input: water_GMX50_bare Ns Per Day > Higher Is Better m7g.16xlarge Graviton3 ... 4.223 |======================================= c6g.16xlarge Graviton2 ... 2.767 |========================== c7g.16xlarge Graviton3 ... 4.200 |======================================= c7gn.16xlarge Graviton3E . 4.820 |============================================= c6a.16xlarge AMD Zen 3 ... 3.965 |===================================== egeo-07 .................. 1.116 |========== Rodinia 3.1 Test: OpenMP CFD Solver Seconds < Lower Is Better m7g.16xlarge Graviton3 ... 4.375 |========== c6g.16xlarge Graviton2 ... 6.051 |============== c7g.16xlarge Graviton3 ... 4.442 |========== c7gn.16xlarge Graviton3E . 4.429 |========== c6a.16xlarge AMD Zen 3 ... 9.342 |====================== egeo-07 .................. 18.630 |============================================ 7-Zip Compression 22.01 Test: Compression Rating MIPS > Higher Is Better m7g.16xlarge Graviton3 ... 316825 |============================================ c6g.16xlarge Graviton2 ... 240702 |================================= c7g.16xlarge Graviton3 ... 311056 |=========================================== c7gn.16xlarge Graviton3E . 312009 |=========================================== c6a.16xlarge AMD Zen 3 ... 230970 |================================ egeo-07 .................. 75840 |=========== Liquid-DSP 1.6 Threads: 32 - Buffer Length: 256 - Filter Length: 512 samples/s > Higher Is Better m7g.16xlarge Graviton3 ... 81396667 |============ c6g.16xlarge Graviton2 ... 67486333 |========== c7g.16xlarge Graviton3 ... 81412000 |============ c7gn.16xlarge Graviton3E . 81394000 |============ c6a.16xlarge AMD Zen 3 ... 274803333 |========================================= egeo-07 .................. 128250000 |=================== Algebraic Multi-Grid Benchmark 1.2 Figure Of Merit > Higher Is Better m7g.16xlarge Graviton3 ... 1646761667 |===================================== c6g.16xlarge Graviton2 ... 1035586333 |======================= c7g.16xlarge Graviton3 ... 1765277667 |======================================== c7gn.16xlarge Graviton3E . 1765966333 |======================================== c6a.16xlarge AMD Zen 3 ... 836999300 |=================== egeo-07 .................. 444456133 |========== OpenSSL 3.1 Algorithm: RSA4096 sign/s > Higher Is Better m7g.16xlarge Graviton3 ... 10181.9 |=========================================== c6g.16xlarge Graviton2 ... 2624.3 |=========== c7g.16xlarge Graviton3 ... 10181.4 |=========================================== c7gn.16xlarge Graviton3E . 10183.3 |=========================================== c6a.16xlarge AMD Zen 3 ... 8392.4 |=================================== egeo-07 .................. 3041.8 |============= Stress-NG 0.15.10 Test: Wide Vector Math Bogo Ops/s > Higher Is Better m7g.16xlarge Graviton3 ... 1542834.94 |======================================== c6g.16xlarge Graviton2 ... 997272.65 |========================== c7g.16xlarge Graviton3 ... 1535336.57 |======================================== c7gn.16xlarge Graviton3E . 1530043.52 |======================================== c6a.16xlarge AMD Zen 3 ... 1380146.63 |==================================== egeo-07 .................. 399851.85 |========== Laghos 3.1 Test: Triple Point Problem Major Kernels Total Rate > Higher Is Better m7g.16xlarge Graviton3 ... 232.01 |=========================================== c6g.16xlarge Graviton2 ... 180.80 |================================== c7g.16xlarge Graviton3 ... 230.68 |=========================================== c7gn.16xlarge Graviton3E . 236.22 |============================================ c6a.16xlarge AMD Zen 3 ... 227.40 |========================================== egeo-07 .................. 61.58 |=========== nginx 1.23.2 Connections: 1000 Requests Per Second > Higher Is Better m7g.16xlarge Graviton3 ... 255616.04 |========================================= c6g.16xlarge Graviton2 ... 158676.40 |========================= c7g.16xlarge Graviton3 ... 255552.05 |========================================= c7gn.16xlarge Graviton3E . 256585.83 |========================================= c6a.16xlarge AMD Zen 3 ... 163178.67 |========================== egeo-07 .................. 70348.77 |=========== QMCPACK 3.16 Input: Li2_STO_ae Total Execution Time - Seconds < Lower Is Better m7g.16xlarge Graviton3 ... 112.61 |============ c6g.16xlarge Graviton2 ... 165.12 |================== c7g.16xlarge Graviton3 ... 112.64 |============ c7gn.16xlarge Graviton3E . 113.20 |============ c6a.16xlarge AMD Zen 3 ... 123.95 |============= egeo-07 .................. 408.45 |============================================ Liquid-DSP 1.6 Threads: 64 - Buffer Length: 256 - Filter Length: 57 samples/s > Higher Is Better m7g.16xlarge Graviton3 ... 1442400000 |================================== c6g.16xlarge Graviton2 ... 978200000 |======================= c7g.16xlarge Graviton3 ... 1442366667 |================================== c7gn.16xlarge Graviton3E . 1442666667 |================================== c6a.16xlarge AMD Zen 3 ... 1710800000 |======================================== egeo-07 .................. 477896667 |=========== OpenSSL 3.1 Algorithm: RSA4096 verify/s > Higher Is Better m7g.16xlarge Graviton3 ... 713859.5 |========================================== c6g.16xlarge Graviton2 ... 214040.9 |============= c7g.16xlarge Graviton3 ... 713945.9 |========================================== c7gn.16xlarge Graviton3E . 713754.8 |========================================== c6a.16xlarge AMD Zen 3 ... 548396.5 |================================ egeo-07 .................. 200342.7 |============ Liquid-DSP 1.6 Threads: 64 - Buffer Length: 256 - Filter Length: 512 samples/s > Higher Is Better m7g.16xlarge Graviton3 ... 162753333 |=============== c6g.16xlarge Graviton2 ... 134926667 |============ c7g.16xlarge Graviton3 ... 162766667 |=============== c7gn.16xlarge Graviton3E . 162756667 |=============== c6a.16xlarge AMD Zen 3 ... 460076667 |========================================= egeo-07 .................. 129560000 |============ Liquid-DSP 1.6 Threads: 64 - Buffer Length: 256 - Filter Length: 32 samples/s > Higher Is Better m7g.16xlarge Graviton3 ... 2270500000 |======================================== c6g.16xlarge Graviton2 ... 1531400000 |=========================== c7g.16xlarge Graviton3 ... 2271966667 |======================================== c7gn.16xlarge Graviton3E . 2266833333 |======================================== c6a.16xlarge AMD Zen 3 ... 2184866667 |====================================== egeo-07 .................. 641243333 |=========== srsRAN Project 23.5 Test: Downlink Processor Benchmark Mbps > Higher Is Better m7g.16xlarge Graviton3 ... 318.5 |===================== c6g.16xlarge Graviton2 ... 197.2 |============= c7g.16xlarge Graviton3 ... 319.7 |===================== c7gn.16xlarge Graviton3E . 323.2 |===================== c6a.16xlarge AMD Zen 3 ... 691.3 |============================================= egeo-07 .................. 342.3 |====================== NAS Parallel Benchmarks 3.4 Test / Class: SP.C Total Mop/s > Higher Is Better m7g.16xlarge Graviton3 ... 17244.85 |===================== c6g.16xlarge Graviton2 ... 9711.70 |============ c7g.16xlarge Graviton3 ... 17219.95 |===================== c7gn.16xlarge Graviton3E . 17163.11 |===================== c6a.16xlarge AMD Zen 3 ... 34025.35 |========================================== egeo-07 .................. 11726.55 |============== OpenSSL 3.1 Algorithm: ChaCha20-Poly1305 byte/s > Higher Is Better m7g.16xlarge Graviton3 ... 74287460990 |=============================== c6g.16xlarge Graviton2 ... 46717636807 |==================== c7g.16xlarge Graviton3 ... 74318842213 |=============================== c7gn.16xlarge Graviton3E . 79969465487 |================================== c6a.16xlarge AMD Zen 3 ... 92522999373 |======================================= egeo-07 .................. 26507041907 |=========== nginx 1.23.2 Connections: 500 Requests Per Second > Higher Is Better m7g.16xlarge Graviton3 ... 255768.44 |========================================= c6g.16xlarge Graviton2 ... 148964.69 |======================== c7g.16xlarge Graviton3 ... 255145.52 |========================================= c7gn.16xlarge Graviton3E . 253518.51 |========================================= c6a.16xlarge AMD Zen 3 ... 165847.75 |=========================== egeo-07 .................. 73654.19 |============ Monte Carlo Simulations of Ionised Nebulae 2.02.73.3 Input: Dust 2D tau100.0 Seconds < Lower Is Better m7g.16xlarge Graviton3 ... 82.67 |============= c6g.16xlarge Graviton2 ... 145.37 |======================= c7g.16xlarge Graviton3 ... 82.82 |============= c7gn.16xlarge Graviton3E . 82.97 |============= c6a.16xlarge AMD Zen 3 ... 194.44 |============================== egeo-07 .................. 281.25 |============================================ srsRAN Project 23.5 Test: PUSCH Processor Benchmark, Throughput Thread Mbps > Higher Is Better m7g.16xlarge Graviton3 ... 95.8 |==================== c6g.16xlarge Graviton2 ... 63.8 |============= c7g.16xlarge Graviton3 ... 95.7 |==================== c7gn.16xlarge Graviton3E . 97.4 |==================== c6a.16xlarge AMD Zen 3 ... 215.9 |============================================= egeo-07 .................. 88.5 |================== Liquid-DSP 1.6 Threads: 32 - Buffer Length: 256 - Filter Length: 57 samples/s > Higher Is Better m7g.16xlarge Graviton3 ... 721493333 |==================== c6g.16xlarge Graviton2 ... 489270000 |============== c7g.16xlarge Graviton3 ... 721386667 |==================== c7gn.16xlarge Graviton3E . 721380000 |==================== c6a.16xlarge AMD Zen 3 ... 1444266667 |======================================== egeo-07 .................. 429596667 |============ QMCPACK 3.16 Input: FeCO6_b3lyp_gms Total Execution Time - Seconds < Lower Is Better m7g.16xlarge Graviton3 ... 211.60 |=============== c6g.16xlarge Graviton2 ... 302.19 |====================== c7g.16xlarge Graviton3 ... 211.32 |=============== c7gn.16xlarge Graviton3E . 188.28 |============== c6a.16xlarge AMD Zen 3 ... 184.10 |============= egeo-07 .................. 608.86 |============================================ Kripke 1.2.6 Throughput FoM > Higher Is Better m7g.16xlarge Graviton3 ... 339000400 |======================================= c6g.16xlarge Graviton2 ... 220120233 |========================= c7g.16xlarge Graviton3 ... 354442733 |========================================= c7gn.16xlarge Graviton3E . 354234067 |========================================= c6a.16xlarge AMD Zen 3 ... 237087650 |=========================== egeo-07 .................. 109107233 |============= QMCPACK 3.16 Input: FeCO6_b3lyp_gms Total Execution Time - Seconds < Lower Is Better m7g.16xlarge Graviton3 ... 205.72 |================ c6g.16xlarge Graviton2 ... 297.94 |======================== c7g.16xlarge Graviton3 ... 204.77 |================ c7gn.16xlarge Graviton3E . 204.25 |================ c6a.16xlarge AMD Zen 3 ... 187.32 |=============== egeo-07 .................. 552.72 |============================================ QMCPACK 3.16 Input: simple-H2O Total Execution Time - Seconds < Lower Is Better m7g.16xlarge Graviton3 ... 28.04 |================ c6g.16xlarge Graviton2 ... 45.23 |========================== c7g.16xlarge Graviton3 ... 27.99 |================ c7gn.16xlarge Graviton3E . 28.00 |================ c6a.16xlarge AMD Zen 3 ... 26.87 |================ egeo-07 .................. 77.52 |============================================= NAS Parallel Benchmarks 3.4 Test / Class: EP.D Total Mop/s > Higher Is Better m7g.16xlarge Graviton3 ... 3738.98 |=========================================== c6g.16xlarge Graviton2 ... 2216.26 |========================= c7g.16xlarge Graviton3 ... 3664.54 |========================================== c7gn.16xlarge Graviton3E . 3657.67 |========================================== c6a.16xlarge AMD Zen 3 ... 3061.42 |=================================== egeo-07 .................. 1329.20 |=============== Timed Node.js Compilation 19.8.1 Time To Compile Seconds < Lower Is Better m7g.16xlarge Graviton3 ... 237.78 |================ c6g.16xlarge Graviton2 ... 287.81 |==================== c7g.16xlarge Graviton3 ... 238.54 |================ c7gn.16xlarge Graviton3E . 238.64 |================ c6a.16xlarge AMD Zen 3 ... 230.42 |================ egeo-07 .................. 641.87 |============================================ Rodinia 3.1 Test: OpenMP Streamcluster Seconds < Lower Is Better m7g.16xlarge Graviton3 ... 11.663 |====================== c6g.16xlarge Graviton2 ... 13.735 |========================== c7g.16xlarge Graviton3 ... 11.625 |====================== c7gn.16xlarge Graviton3E . 10.690 |==================== c6a.16xlarge AMD Zen 3 ... 8.396 |================ egeo-07 .................. 23.247 |============================================ Stress-NG 0.15.10 Test: CPU Cache Bogo Ops/s > Higher Is Better m7g.16xlarge Graviton3 ... 3892396.34 |======================================== c6g.16xlarge Graviton2 ... 1921785.20 |==================== c7g.16xlarge Graviton3 ... 3844101.98 |======================================== c7gn.16xlarge Graviton3E . 3860335.38 |======================================== c6a.16xlarge AMD Zen 3 ... 1447265.35 |=============== egeo-07 .................. 1525522.46 |================ Timed Godot Game Engine Compilation 4.0 Time To Compile Seconds < Lower Is Better m7g.16xlarge Graviton3 ... 154.38 |================= c6g.16xlarge Graviton2 ... 218.28 |========================= c7g.16xlarge Graviton3 ... 156.69 |================== c7gn.16xlarge Graviton3E . 155.95 |================== c6a.16xlarge AMD Zen 3 ... 147.74 |================= egeo-07 .................. 391.74 |============================================ NAS Parallel Benchmarks 3.4 Test / Class: MG.C Total Mop/s > Higher Is Better m7g.16xlarge Graviton3 ... 50126.29 |========================================== c6g.16xlarge Graviton2 ... 25671.29 |====================== c7g.16xlarge Graviton3 ... 49742.30 |========================================== c7gn.16xlarge Graviton3E . 49860.68 |========================================== c6a.16xlarge AMD Zen 3 ... 45946.81 |====================================== egeo-07 .................. 19477.23 |================ Timed Gem5 Compilation 21.2 Time To Compile Seconds < Lower Is Better m7g.16xlarge Graviton3 ... 180.25 |================= c6g.16xlarge Graviton2 ... 225.31 |===================== c7g.16xlarge Graviton3 ... 181.78 |================= c7gn.16xlarge Graviton3E . 182.47 |================= c6a.16xlarge AMD Zen 3 ... 192.12 |================== egeo-07 .................. 462.60 |============================================ OpenSSL 3.1 Algorithm: ChaCha20 byte/s > Higher Is Better m7g.16xlarge Graviton3 ... 103226784517 |============================ c6g.16xlarge Graviton2 ... 67292541203 |================== c7g.16xlarge Graviton3 ... 103275516997 |============================ c7gn.16xlarge Graviton3E . 114118119423 |=============================== c6a.16xlarge AMD Zen 3 ... 138389378753 |====================================== egeo-07 .................. 56213494627 |=============== nekRS 23.0 Input: Kershaw flops/rank > Higher Is Better m7g.16xlarge Graviton3 ... 3150680000 |============================= c6g.16xlarge Graviton2 ... 1760336667 |================ c7g.16xlarge Graviton3 ... 3261853333 |============================== c7gn.16xlarge Graviton3E . 3302823333 |=============================== c6a.16xlarge AMD Zen 3 ... 4308810000 |======================================== NAS Parallel Benchmarks 3.4 Test / Class: CG.C Total Mop/s > Higher Is Better m7g.16xlarge Graviton3 ... 21988.99 |========================================== c6g.16xlarge Graviton2 ... 13103.62 |========================= c7g.16xlarge Graviton3 ... 21911.02 |========================================== c7gn.16xlarge Graviton3E . 22155.36 |========================================== c6a.16xlarge AMD Zen 3 ... 20210.00 |====================================== egeo-07 .................. 9661.91 |================== Monte Carlo Simulations of Ionised Nebulae 2.02.73.3 Input: Gas HII40 Seconds < Lower Is Better m7g.16xlarge Graviton3 ... 13.58 |======================= c6g.16xlarge Graviton2 ... 20.76 |=================================== c7g.16xlarge Graviton3 ... 13.66 |======================= c7gn.16xlarge Graviton3E . 13.53 |======================= c6a.16xlarge AMD Zen 3 ... 12.67 |===================== egeo-07 .................. 26.92 |============================================= OpenSSL 3.1 Algorithm: SHA256 byte/s > Higher Is Better m7g.16xlarge Graviton3 ... 54212515580 |======================================= c6g.16xlarge Graviton2 ... 42472798847 |=============================== c7g.16xlarge Graviton3 ... 54216561263 |======================================= c7gn.16xlarge Graviton3E . 54154218593 |======================================= c6a.16xlarge AMD Zen 3 ... 45857534777 |================================= egeo-07 .................. 3466925203 |== ACES DGEMM 1.0 Sustained Floating-Point Rate GFLOP/s > Higher Is Better m7g.16xlarge Graviton3 ... 24.362353 |========================================= c6g.16xlarge Graviton2 ... 20.417952 |================================== c7g.16xlarge Graviton3 ... 24.140605 |========================================= c7gn.16xlarge Graviton3E . 24.078529 |========================================= c6a.16xlarge AMD Zen 3 ... 9.388050 |================ egeo-07 .................. 2.094206 |==== nekRS 23.0 Input: TurboPipe Periodic flops/rank > Higher Is Better m7g.16xlarge Graviton3 ... 3976300000 |===================================== c6g.16xlarge Graviton2 ... 2220190000 |==================== c7g.16xlarge Graviton3 ... 3978983333 |===================================== c7gn.16xlarge Graviton3E . 4141440000 |====================================== c6a.16xlarge AMD Zen 3 ... 4337536667 |======================================== Liquid-DSP 1.6 Threads: 32 - Buffer Length: 256 - Filter Length: 32 samples/s > Higher Is Better m7g.16xlarge Graviton3 ... 1136066667 |====================================== c6g.16xlarge Graviton2 ... 765466667 |========================== c7g.16xlarge Graviton3 ... 1136133333 |====================================== c7gn.16xlarge Graviton3E . 1136000000 |====================================== c6a.16xlarge AMD Zen 3 ... 1193966667 |======================================== egeo-07 .................. 633483333 |===================== LeelaChessZero 0.28 Backend: Eigen Nodes Per Second > Higher Is Better m7g.16xlarge Graviton3 ... 1398 |============================================= c6g.16xlarge Graviton2 ... 891 |============================ c7g.16xlarge Graviton3 ... 1382 |============================================ c7gn.16xlarge Graviton3E . 1444 |============================================== c6a.16xlarge AMD Zen 3 ... 1152 |===================================== LeelaChessZero 0.28 Backend: BLAS Nodes Per Second > Higher Is Better m7g.16xlarge Graviton3 ... 1301 |=========================================== c6g.16xlarge Graviton2 ... 947 |=============================== c7g.16xlarge Graviton3 ... 1333 |============================================ c7gn.16xlarge Graviton3E . 1392 |============================================== c6a.16xlarge AMD Zen 3 ... 1316 |=========================================== Stockfish 15 Total Time Nodes Per Second > Higher Is Better m7g.16xlarge Graviton3 ... 112119711 |======================================= c6g.16xlarge Graviton2 ... 86609284 |============================== c7g.16xlarge Graviton3 ... 117316476 |========================================= c7gn.16xlarge Graviton3E . 117027121 |========================================= c6a.16xlarge AMD Zen 3 ... 96905609 |================================== egeo-07 .................. 26092344 |=========