Benchmarks by Michael Larabel for a future article looking at AWS Graviton1, Graviton2, Graviton3, Graviton4 metal performance.
Graviton1 16 Cores a1.metal Processor: ARMv8 Cortex-A72 (16 Cores), Motherboard: Amazon EC2 a1.metal (1.0 BIOS), Memory: 32GB, Disk: 429GB Amazon Elastic Block Store, Network: Amazon Elastic
OS: Ubuntu 24.04, Kernel: 6.8.0-1009-aws (aarch64), Compiler: GCC 13.2.0, File-System: ext4
Kernel Notes: Transparent Huge Pages: madviseCompiler Notes: --build=aarch64-linux-gnu --disable-libquadmath --disable-libquadmath-support --disable-werror --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-fix-cortex-a53-843419 --enable-gnu-unique-object --enable-languages=c,ada,c++,go,d,fortran,objc,obj-c++,m2 --enable-libphobos-checking=release --enable-libstdcxx-backtrace --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-multiarch --enable-nls --enable-objc-gc=auto --enable-offload-defaulted --enable-offload-targets=nvptx-none=/build/gcc-13-dIwDw0/gcc-13-13.2.0/debian/tmp-nvptx/usr --enable-plugin --enable-shared --enable-threads=posix --host=aarch64-linux-gnu --program-prefix=aarch64-linux-gnu- --target=aarch64-linux-gnu --with-default-libstdcxx-abi=new --with-gcc-major-version-only --with-target-system-zlib=auto --without-cuda-driver -vJava Notes: OpenJDK Runtime Environment (build 11.0.23+9-post-Ubuntu-1ubuntu1)Python Notes: Python 3.12.3Security Notes: gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + reg_file_data_sampling: Not affected + retbleed: Not affected + spec_rstack_overflow: Not affected + spec_store_bypass: Not affected + spectre_v1: Mitigation of __user pointer sanitization + spectre_v2: Mitigation of Branch predictor hardening BHB + srbds: Not affected + tsx_async_abort: Not affected
Graviton2 64 Cores r6g.metal Changed Processor to ARMv8 Neoverse-N1 (64 Cores) .
Changed Motherboard to Amazon EC2 r6g.metal (1.0 BIOS) .
Changed Memory to 8 x 64GB DDR4-3200MT/s .
Security Change: gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + reg_file_data_sampling: Not affected + retbleed: Not affected + spec_rstack_overflow: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of __user pointer sanitization + spectre_v2: Mitigation of CSV2 BHB + srbds: Not affected + tsx_async_abort: Not affected
Graviton3 64 Cores r7g.metal Changed Processor to ARMv8 Neoverse-V1 (64 Cores) .
Changed Motherboard to Amazon EC2 r7g.metal (1.0 BIOS) .
Changed Memory to 8 x 64GB DDR5-4400MT/s .
Graviton4 96 Cores r8g.metal-24xl Changed Processor to ARMv8 Neoverse-V2 (96 Cores) .
Changed Motherboard to Amazon EC2 r8g.metal-24xl (1.0 BIOS) .
Changed Memory to 12 x 64GB DDR5-5600MT/s .
Security Change: gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + reg_file_data_sampling: Not affected + retbleed: Not affected + spec_rstack_overflow: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of __user pointer sanitization + spectre_v2: Not affected + srbds: Not affected + tsx_async_abort: Not affected
Timed Node.js Compilation This test profile times how long it takes to build/compile Node.js itself from source. Node.js is a JavaScript run-time built from the Chrome V8 JavaScript engine while itself is written in C/C++. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Seconds, Fewer Is Better Timed Node.js Compilation 21.7.2 Time To Compile Graviton4 96 Cores r8g.metal-24xl Graviton3 64 Cores r7g.metal Graviton2 64 Cores r6g.metal Graviton1 16 Cores a1.metal 1000 2000 3000 4000 5000 SE +/- 0.69, N = 3 SE +/- 0.73, N = 3 SE +/- 0.21, N = 3 SE +/- 1.44, N = 3 289.27 555.67 647.62 4663.28
Stockfish This is a test of Stockfish, an advanced open-source C++11 chess benchmark that can scale up to 1024 CPU threads. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Nodes Per Second, More Is Better Stockfish 16.1 Chess Benchmark Graviton4 96 Cores r8g.metal-24xl Graviton3 64 Cores r7g.metal Graviton2 64 Cores r6g.metal Graviton1 16 Cores a1.metal 30M 60M 90M 120M 150M SE +/- 2918876.82, N = 15 SE +/- 1588787.71, N = 12 SE +/- 1121445.57, N = 12 SE +/- 39945.87, N = 9 118337918 57635786 43949837 2493057 1. (CXX) g++ options: -lgcov -lpthread -fno-exceptions -std=c++17 -fno-peel-loops -fno-tracer -pedantic -O3 -funroll-loops -flto -flto-partition=one -flto=jobserver
Timed Godot Game Engine Compilation This test times how long it takes to compile the Godot Game Engine. Godot is a popular, open-source, cross-platform 2D/3D game engine and is built using the SCons build system and targeting the X11 platform. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Seconds, Fewer Is Better Timed Godot Game Engine Compilation 4.0 Time To Compile Graviton4 96 Cores r8g.metal-24xl Graviton3 64 Cores r7g.metal Graviton2 64 Cores r6g.metal Graviton1 16 Cores a1.metal 400 800 1200 1600 2000 SE +/- 1.14, N = 3 SE +/- 0.44, N = 3 SE +/- 0.92, N = 3 SE +/- 0.31, N = 3 126.35 190.35 266.14 1674.47
Blender Blender is an open-source 3D creation software project. This test is of Blender's Cycles benchmark with various sample files. GPU computing is supported. This system/blender test profile makes use of the system-supplied Blender. Use pts/blender if wishing to stick to a fixed version of Blender. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Seconds, Fewer Is Better Blender 4.0.2 Blend File: Classroom - Compute: CPU-Only Graviton4 96 Cores r8g.metal-24xl Graviton3 64 Cores r7g.metal Graviton2 64 Cores r6g.metal Graviton1 16 Cores a1.metal 300 600 900 1200 1500 SE +/- 0.03, N = 3 SE +/- 0.09, N = 3 SE +/- 0.35, N = 3 SE +/- 0.28, N = 3 69.40 128.36 166.70 1289.61
Blender Blender is an open-source 3D creation software project. This test is of Blender's Cycles benchmark with various sample files. GPU computing is supported. This system/blender test profile makes use of the system-supplied Blender. Use pts/blender if wishing to stick to a fixed version of Blender. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Seconds, Fewer Is Better Blender 4.0.2 Blend File: Fishy Cat - Compute: CPU-Only Graviton4 96 Cores r8g.metal-24xl Graviton3 64 Cores r7g.metal Graviton2 64 Cores r6g.metal Graviton1 16 Cores a1.metal 200 400 600 800 1000 SE +/- 0.26, N = 3 SE +/- 0.29, N = 3 SE +/- 0.35, N = 3 SE +/- 0.38, N = 3 64.07 114.60 146.68 952.43
C-Ray OpenBenchmarking.org Seconds, Fewer Is Better C-Ray 2.0 Resolution: 5K - Rays Per Pixel: 16 Graviton4 96 Cores r8g.metal-24xl Graviton3 64 Cores r7g.metal Graviton2 64 Cores r6g.metal Graviton1 16 Cores a1.metal 200 400 600 800 1000 SE +/- 0.03, N = 3 SE +/- 0.06, N = 3 SE +/- 0.03, N = 3 SE +/- 2.98, N = 3 33.68 68.58 112.87 1008.69 1. (CC) gcc options: -lpthread -lm
libavif avifenc This is a test of the AOMedia libavif library testing the encoding of a JPEG image to AV1 Image Format (AVIF). Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Seconds, Fewer Is Better libavif avifenc 1.0 Encoder Speed: 0 Graviton4 96 Cores r8g.metal-24xl Graviton3 64 Cores r7g.metal Graviton2 64 Cores r6g.metal Graviton1 16 Cores a1.metal 140 280 420 560 700 SE +/- 0.01, N = 3 SE +/- 0.07, N = 3 SE +/- 0.17, N = 3 SE +/- 3.02, N = 3 130.20 157.28 251.33 668.50 1. (CXX) g++ options: -O3 -fPIC -lm
PostgreSQL This is a benchmark of PostgreSQL using the integrated pgbench for facilitating the database benchmarks. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org ms, Fewer Is Better PostgreSQL 16 Scaling Factor: 100 - Clients: 1000 - Mode: Read Only - Average Latency Graviton4 96 Cores r8g.metal-24xl Graviton3 64 Cores r7g.metal Graviton2 64 Cores r6g.metal Graviton1 16 Cores a1.metal 3 6 9 12 15 SE +/- 0.002, N = 3 SE +/- 0.003, N = 3 SE +/- 0.008, N = 10 SE +/- 0.010, N = 3 0.346 0.867 1.015 9.816 1. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm
OpenBenchmarking.org TPS, More Is Better PostgreSQL 16 Scaling Factor: 100 - Clients: 1000 - Mode: Read Only Graviton4 96 Cores r8g.metal-24xl Graviton3 64 Cores r7g.metal Graviton2 64 Cores r6g.metal Graviton1 16 Cores a1.metal 600K 1200K 1800K 2400K 3000K SE +/- 15759.90, N = 3 SE +/- 3990.43, N = 3 SE +/- 7623.84, N = 10 SE +/- 107.42, N = 3 2893264 1153677 985509 101870 1. (CC) gcc options: -fno-strict-aliasing -fwrapv -O2 -lpgcommon -lpgport -lpq -lm
High Performance Conjugate Gradient HPCG is the High Performance Conjugate Gradient and is a new scientific benchmark from Sandia National Lans focused for super-computer testing with modern real-world workloads compared to HPCC. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org GFLOP/s, More Is Better High Performance Conjugate Gradient 3.1 X Y Z: 104 104 104 - RT: 60 Graviton4 96 Cores r8g.metal-24xl Graviton3 64 Cores r7g.metal Graviton2 64 Cores r6g.metal Graviton1 16 Cores a1.metal 14 28 42 56 70 SE +/- 0.01489, N = 3 SE +/- 0.01313, N = 3 SE +/- 0.00037, N = 3 SE +/- 0.00496, N = 3 61.61950 33.07990 20.55910 3.76867 1. (CXX) g++ options: -O3 -ffast-math -ftree-vectorize -lmpi_cxx -lmpi
Speedb Speedb is a next-generation key value storage engine that is RocksDB compatible and aiming for stability, efficiency, and performance. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Op/s, More Is Better Speedb 2.7 Test: Read While Writing Graviton4 96 Cores r8g.metal-24xl Graviton3 64 Cores r7g.metal Graviton2 64 Cores r6g.metal Graviton1 16 Cores a1.metal 2M 4M 6M 8M 10M SE +/- 87574.70, N = 15 SE +/- 23497.61, N = 3 SE +/- 74542.67, N = 15 SE +/- 20302.77, N = 8 11490911 8758376 8229780 2433641 1. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti
Blender Blender is an open-source 3D creation software project. This test is of Blender's Cycles benchmark with various sample files. GPU computing is supported. This system/blender test profile makes use of the system-supplied Blender. Use pts/blender if wishing to stick to a fixed version of Blender. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Seconds, Fewer Is Better Blender 4.0.2 Blend File: BMW27 - Compute: CPU-Only Graviton4 96 Cores r8g.metal-24xl Graviton3 64 Cores r7g.metal Graviton2 64 Cores r6g.metal Graviton1 16 Cores a1.metal 130 260 390 520 650 SE +/- 0.03, N = 3 SE +/- 0.18, N = 3 SE +/- 0.21, N = 3 SE +/- 0.63, N = 3 33.72 62.63 81.54 599.39
OpenSSL OpenBenchmarking.org byte/s, More Is Better OpenSSL Algorithm: ChaCha20 Graviton4 96 Cores r8g.metal-24xl Graviton3 64 Cores r7g.metal Graviton2 64 Cores r6g.metal Graviton1 16 Cores a1.metal 30000M 60000M 90000M 120000M 150000M SE +/- 430328.22, N = 3 SE +/- 22981194.52, N = 3 SE +/- 157650.92, N = 3 SE +/- 4780384.74, N = 3 148791688760 94241563073 67238751070 8285450333 1. OpenSSL 3.0.13 30 Jan 2024 (Library: OpenSSL 3.0.13 30 Jan 2024)
OpenBenchmarking.org byte/s, More Is Better OpenSSL Algorithm: AES-128-GCM Graviton4 96 Cores r8g.metal-24xl Graviton3 64 Cores r7g.metal Graviton2 64 Cores r6g.metal Graviton1 16 Cores a1.metal 100000M 200000M 300000M 400000M 500000M SE +/- 2173107.96, N = 3 SE +/- 2741312.53, N = 3 SE +/- 27172495.86, N = 3 SE +/- 35254075.37, N = 3 444871148797 288131192333 157386309813 30455741337 1. OpenSSL 3.0.13 30 Jan 2024 (Library: OpenSSL 3.0.13 30 Jan 2024)
OpenBenchmarking.org byte/s, More Is Better OpenSSL Algorithm: SHA256 Graviton4 96 Cores r8g.metal-24xl Graviton3 64 Cores r7g.metal Graviton2 64 Cores r6g.metal Graviton1 16 Cores a1.metal 20000M 40000M 60000M 80000M 100000M SE +/- 7443314.27, N = 3 SE +/- 21492664.43, N = 3 SE +/- 308447238.10, N = 3 SE +/- 62991235.00, N = 3 85604846770 47482413893 38976308003 6151749933 1. OpenSSL 3.0.13 30 Jan 2024 (Library: OpenSSL 3.0.13 30 Jan 2024)
OpenBenchmarking.org byte/s, More Is Better OpenSSL Algorithm: ChaCha20-Poly1305 Graviton4 96 Cores r8g.metal-24xl Graviton3 64 Cores r7g.metal Graviton2 64 Cores r6g.metal Graviton1 16 Cores a1.metal 20000M 40000M 60000M 80000M 100000M SE +/- 512363.59, N = 3 SE +/- 7386579.63, N = 3 SE +/- 281207.00, N = 3 SE +/- 4948671.03, N = 3 110699450003 69534325600 46636998930 6554146770 1. OpenSSL 3.0.13 30 Jan 2024 (Library: OpenSSL 3.0.13 30 Jan 2024)
OpenBenchmarking.org byte/s, More Is Better OpenSSL Algorithm: AES-256-GCM Graviton4 96 Cores r8g.metal-24xl Graviton3 64 Cores r7g.metal Graviton2 64 Cores r6g.metal Graviton1 16 Cores a1.metal 90000M 180000M 270000M 360000M 450000M SE +/- 15611507.33, N = 3 SE +/- 48618733.94, N = 3 SE +/- 5879487.16, N = 3 SE +/- 17186337.18, N = 3 400130558450 250606646610 126846323187 25773496457 1. OpenSSL 3.0.13 30 Jan 2024 (Library: OpenSSL 3.0.13 30 Jan 2024)
OpenBenchmarking.org byte/s, More Is Better OpenSSL Algorithm: SHA512 Graviton4 96 Cores r8g.metal-24xl Graviton3 64 Cores r7g.metal Graviton2 64 Cores r6g.metal Graviton1 16 Cores a1.metal 11000M 22000M 33000M 44000M 55000M SE +/- 4634061.90, N = 3 SE +/- 2215754.50, N = 3 SE +/- 20829920.25, N = 3 SE +/- 4619831.10, N = 3 53203418283 30714577477 14122869463 2505891877 1. OpenSSL 3.0.13 30 Jan 2024 (Library: OpenSSL 3.0.13 30 Jan 2024)
GROMACS The GROMACS (GROningen MAchine for Chemical Simulations) molecular dynamics package testing with the water_GMX50 data. This test profile allows selecting between CPU and GPU-based GROMACS builds. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Ns Per Day, More Is Better GROMACS 2024 Implementation: MPI CPU - Input: water_GMX50_bare Graviton4 96 Cores r8g.metal-24xl Graviton3 64 Cores r7g.metal Graviton2 64 Cores r6g.metal Graviton1 16 Cores a1.metal 2 4 6 8 10 SE +/- 0.001, N = 3 SE +/- 0.002, N = 3 SE +/- 0.001, N = 3 SE +/- 0.000, N = 3 6.926 4.206 2.731 0.313 1. (CXX) g++ options: -O3 -lm
libavif avifenc This is a test of the AOMedia libavif library testing the encoding of a JPEG image to AV1 Image Format (AVIF). Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Seconds, Fewer Is Better libavif avifenc 1.0 Encoder Speed: 2 Graviton4 96 Cores r8g.metal-24xl Graviton3 64 Cores r7g.metal Graviton2 64 Cores r6g.metal Graviton1 16 Cores a1.metal 80 160 240 320 400 SE +/- 0.02, N = 3 SE +/- 0.05, N = 3 SE +/- 0.08, N = 3 SE +/- 0.85, N = 3 80.24 95.40 159.51 381.11 1. (CXX) g++ options: -O3 -fPIC -lm
PyPerformance OpenBenchmarking.org Milliseconds, Fewer Is Better PyPerformance 1.11 Benchmark: xml_etree Graviton4 96 Cores r8g.metal-24xl Graviton3 64 Cores r7g.metal Graviton2 64 Cores r6g.metal Graviton1 16 Cores a1.metal 50 100 150 200 250 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.58, N = 3 58.2 76.6 100.0 230.0
C-Ray OpenBenchmarking.org Seconds, Fewer Is Better C-Ray 2.0 Resolution: 4K - Rays Per Pixel: 16 Graviton4 96 Cores r8g.metal-24xl Graviton3 64 Cores r7g.metal Graviton2 64 Cores r6g.metal Graviton1 16 Cores a1.metal 120 240 360 480 600 SE +/- 0.02, N = 3 SE +/- 0.02, N = 3 SE +/- 0.02, N = 3 SE +/- 1.65, N = 3 18.93 38.56 63.49 566.95 1. (CC) gcc options: -lpthread -lm
RocksDB This is a benchmark of Meta/Facebook's RocksDB as an embeddable persistent key-value store for fast storage based on Google's LevelDB. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Op/s, More Is Better RocksDB 9.0 Test: Update Random Graviton4 96 Cores r8g.metal-24xl Graviton3 64 Cores r7g.metal Graviton2 64 Cores r6g.metal Graviton1 16 Cores a1.metal 300K 600K 900K 1200K 1500K SE +/- 5313.65, N = 3 SE +/- 13144.46, N = 3 SE +/- 4819.65, N = 15 SE +/- 443.99, N = 3 1327423 1026393 728335 164616 1. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti
Speedb Speedb is a next-generation key value storage engine that is RocksDB compatible and aiming for stability, efficiency, and performance. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Op/s, More Is Better Speedb 2.7 Test: Random Read Graviton4 96 Cores r8g.metal-24xl Graviton3 64 Cores r7g.metal Graviton2 64 Cores r6g.metal Graviton1 16 Cores a1.metal 140M 280M 420M 560M 700M SE +/- 229391.69, N = 3 SE +/- 1029086.00, N = 3 SE +/- 445171.45, N = 3 SE +/- 174965.25, N = 15 640938471 284105775 187996683 23286630 1. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti
RocksDB This is a benchmark of Meta/Facebook's RocksDB as an embeddable persistent key-value store for fast storage based on Google's LevelDB. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Op/s, More Is Better RocksDB 9.0 Test: Read While Writing Graviton4 96 Cores r8g.metal-24xl Graviton3 64 Cores r7g.metal Graviton2 64 Cores r6g.metal Graviton1 16 Cores a1.metal 2M 4M 6M 8M 10M SE +/- 18723.10, N = 3 SE +/- 18610.37, N = 3 SE +/- 57133.93, N = 15 SE +/- 16876.08, N = 3 11560819 7164354 6128899 1268650 1. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti
OpenBenchmarking.org Op/s, More Is Better RocksDB 9.0 Test: Random Read Graviton4 96 Cores r8g.metal-24xl Graviton3 64 Cores r7g.metal Graviton2 64 Cores r6g.metal Graviton1 16 Cores a1.metal 110M 220M 330M 440M 550M SE +/- 53499.62, N = 3 SE +/- 86390.65, N = 3 SE +/- 1103933.53, N = 15 SE +/- 12274.71, N = 3 527458547 246989698 165261499 23252821 1. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti
Stress-NG OpenBenchmarking.org Bogo Ops/s, More Is Better Stress-NG 0.17.08 Test: CPU Cache Graviton4 96 Cores r8g.metal-24xl Graviton3 64 Cores r7g.metal Graviton2 64 Cores r6g.metal Graviton1 16 Cores a1.metal 1000K 2000K 3000K 4000K 5000K SE +/- 57877.80, N = 15 SE +/- 45276.44, N = 4 SE +/- 23602.71, N = 15 SE +/- 3895.03, N = 3 4735225.17 3855459.62 2004843.76 631797.13 1. (CXX) g++ options: -O2 -std=gnu99 -lc -lm
PyPerformance OpenBenchmarking.org Milliseconds, Fewer Is Better PyPerformance 1.11 Benchmark: python_startup Graviton4 96 Cores r8g.metal-24xl Graviton3 64 Cores r7g.metal Graviton2 64 Cores r6g.metal Graviton1 16 Cores a1.metal 6 12 18 24 30 SE +/- 0.00, N = 3 SE +/- 0.03, N = 3 SE +/- 0.03, N = 3 SE +/- 0.00, N = 3 7.55 10.10 12.40 25.90
srsRAN Project srsRAN Project is a complete ORAN-native 5G RAN solution created by Software Radio Systems (SRS). The srsRAN Project radio suite was formerly known as srsLTE and can be used for building your own software-defined radio (SDR) 4G/5G mobile network. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Mbps, More Is Better srsRAN Project 23.10.1-20240325 Test: PUSCH Processor Benchmark, Throughput Thread Graviton4 96 Cores r8g.metal-24xl Graviton3 64 Cores r7g.metal Graviton2 64 Cores r6g.metal Graviton1 16 Cores a1.metal 15 30 45 60 75 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.23, N = 4 69.4 53.0 39.4 18.7 MIN: 45.4 MIN: 35.3 MIN: 24.4 MIN: 11.3 / MAX: 18.9 1. (CXX) g++ options: -O3 -fno-trapping-math -fno-math-errno -ldl
ACES DGEMM This is a multi-threaded DGEMM benchmark. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org GFLOP/s, More Is Better ACES DGEMM 1.0 Sustained Floating-Point Rate Graviton4 96 Cores r8g.metal-24xl Graviton3 64 Cores r7g.metal Graviton2 64 Cores r6g.metal Graviton1 16 Cores a1.metal 14 28 42 56 70 SE +/- 0.164173, N = 3 SE +/- 0.005519, N = 3 SE +/- 0.156541, N = 3 SE +/- 0.003641, N = 3 60.635839 34.114458 20.410800 0.879250 1. (CC) gcc options: -O3 -march=native -fopenmp
GraphicsMagick This is a test of GraphicsMagick with its OpenMP implementation that performs various imaging tests on a sample high resolution (currently 15400 x 6940) JPEG image. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Iterations Per Minute, More Is Better GraphicsMagick 1.3.43 Operation: Enhanced Graviton4 96 Cores r8g.metal-24xl Graviton3 64 Cores r7g.metal Graviton2 64 Cores r6g.metal Graviton1 16 Cores a1.metal 60 120 180 240 300 SE +/- 0.67, N = 3 SE +/- 0.00, N = 3 SE +/- 0.33, N = 3 SE +/- 0.00, N = 3 294 195 128 20 1. (CC) gcc options: -fopenmp -O2 -ljpeg -lXext -lSM -lICE -lX11 -lxml2 -lz -lm -lpthread -lgomp
John The Ripper This is a benchmark of John The Ripper, which is a password cracker. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Real C/S, More Is Better John The Ripper 2023.03.14 Test: bcrypt Graviton4 96 Cores r8g.metal-24xl Graviton3 64 Cores r7g.metal Graviton2 64 Cores r6g.metal Graviton1 16 Cores a1.metal 20K 40K 60K 80K 100K SE +/- 22.45, N = 3 SE +/- 1.33, N = 3 SE +/- 0.67, N = 3 SE +/- 94.53, N = 15 85647 50787 45349 7842 1. (CC) gcc options: -lssl -lcrypto -fopenmp -lm -lrt -lz -ldl -lcrypt
Speedb Speedb is a next-generation key value storage engine that is RocksDB compatible and aiming for stability, efficiency, and performance. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Op/s, More Is Better Speedb 2.7 Test: Update Random Graviton4 96 Cores r8g.metal-24xl Graviton3 64 Cores r7g.metal Graviton2 64 Cores r6g.metal Graviton1 16 Cores a1.metal 160K 320K 480K 640K 800K SE +/- 3844.86, N = 3 SE +/- 3804.56, N = 3 SE +/- 3440.64, N = 3 SE +/- 302.33, N = 3 729398 664566 363540 128167 1. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti
John The Ripper This is a benchmark of John The Ripper, which is a password cracker. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Real C/S, More Is Better John The Ripper 2023.03.14 Test: Blowfish Graviton4 96 Cores r8g.metal-24xl Graviton3 64 Cores r7g.metal Graviton2 64 Cores r6g.metal Graviton1 16 Cores a1.metal 20K 40K 60K 80K 100K SE +/- 3.00, N = 3 SE +/- 13.72, N = 3 SE +/- 36.69, N = 3 SE +/- 106.90, N = 15 85667 50774 45305 8003 1. (CC) gcc options: -lssl -lcrypto -fopenmp -lm -lrt -lz -ldl -lcrypt
GraphicsMagick This is a test of GraphicsMagick with its OpenMP implementation that performs various imaging tests on a sample high resolution (currently 15400 x 6940) JPEG image. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Iterations Per Minute, More Is Better GraphicsMagick 1.3.43 Operation: Noise-Gaussian Graviton4 96 Cores r8g.metal-24xl Graviton3 64 Cores r7g.metal Graviton2 64 Cores r6g.metal Graviton1 16 Cores a1.metal 50 100 150 200 250 SE +/- 0.67, N = 3 SE +/- 0.33, N = 3 SE +/- 0.33, N = 3 SE +/- 0.00, N = 3 208 156 117 26 1. (CC) gcc options: -fopenmp -O2 -ljpeg -lXext -lSM -lICE -lX11 -lxml2 -lz -lm -lpthread -lgomp
John The Ripper This is a benchmark of John The Ripper, which is a password cracker. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Real C/S, More Is Better John The Ripper 2023.03.14 Test: HMAC-SHA512 Graviton4 96 Cores r8g.metal-24xl Graviton3 64 Cores r7g.metal Graviton2 64 Cores r6g.metal Graviton1 16 Cores a1.metal 30M 60M 90M 120M 150M SE +/- 162829.50, N = 3 SE +/- 68676.37, N = 3 SE +/- 77846.72, N = 3 SE +/- 9769.57, N = 3 158238333 92007667 64712667 10321667 1. (CC) gcc options: -lssl -lcrypto -fopenmp -lm -lrt -lz -ldl -lcrypt
Speedb Speedb is a next-generation key value storage engine that is RocksDB compatible and aiming for stability, efficiency, and performance. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Op/s, More Is Better Speedb 2.7 Test: Read Random Write Random Graviton4 96 Cores r8g.metal-24xl Graviton3 64 Cores r7g.metal Graviton2 64 Cores r6g.metal Graviton1 16 Cores a1.metal 1.1M 2.2M 3.3M 4.4M 5.5M SE +/- 4241.50, N = 3 SE +/- 3662.12, N = 3 SE +/- 25580.42, N = 3 SE +/- 214.78, N = 3 5342911 3735110 2781037 552621 1. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti
GraphicsMagick This is a test of GraphicsMagick with its OpenMP implementation that performs various imaging tests on a sample high resolution (currently 15400 x 6940) JPEG image. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Iterations Per Minute, More Is Better GraphicsMagick 1.3.43 Operation: Sharpen Graviton4 96 Cores r8g.metal-24xl Graviton3 64 Cores r7g.metal Graviton2 64 Cores r6g.metal Graviton1 16 Cores a1.metal 70 140 210 280 350 SE +/- 0.33, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 325 252 153 26 1. (CC) gcc options: -fopenmp -O2 -ljpeg -lXext -lSM -lICE -lX11 -lxml2 -lz -lm -lpthread -lgomp
John The Ripper This is a benchmark of John The Ripper, which is a password cracker. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Real C/S, More Is Better John The Ripper 2023.03.14 Test: MD5 Graviton4 96 Cores r8g.metal-24xl Graviton3 64 Cores r7g.metal Graviton2 64 Cores r6g.metal Graviton1 16 Cores a1.metal 500K 1000K 1500K 2000K 2500K SE +/- 1201.85, N = 3 SE +/- 1154.70, N = 3 SE +/- 333.33, N = 3 SE +/- 287.69, N = 3 2367333 1453000 1341667 190426 1. (CC) gcc options: -lssl -lcrypto -fopenmp -lm -lrt -lz -ldl -lcrypt
RocksDB This is a benchmark of Meta/Facebook's RocksDB as an embeddable persistent key-value store for fast storage based on Google's LevelDB. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Op/s, More Is Better RocksDB 9.0 Test: Read Random Write Random Graviton4 96 Cores r8g.metal-24xl Graviton3 64 Cores r7g.metal Graviton2 64 Cores r6g.metal Graviton1 16 Cores a1.metal 1.5M 3M 4.5M 6M 7.5M SE +/- 15765.85, N = 3 SE +/- 18849.94, N = 3 SE +/- 5811.82, N = 3 SE +/- 4162.42, N = 3 7047101 4488271 3421023 574458 1. (CXX) g++ options: -O3 -march=native -pthread -fno-builtin-memcmp -fno-rtti
GraphicsMagick This is a test of GraphicsMagick with its OpenMP implementation that performs various imaging tests on a sample high resolution (currently 15400 x 6940) JPEG image. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Iterations Per Minute, More Is Better GraphicsMagick 1.3.43 Operation: HWB Color Space Graviton4 96 Cores r8g.metal-24xl Graviton3 64 Cores r7g.metal Graviton2 64 Cores r6g.metal Graviton1 16 Cores a1.metal 70 140 210 280 350 SE +/- 0.33, N = 3 SE +/- 0.58, N = 3 SE +/- 0.58, N = 3 SE +/- 0.00, N = 3 301 266 212 81 1. (CC) gcc options: -fopenmp -O2 -ljpeg -lXext -lSM -lICE -lX11 -lxml2 -lz -lm -lpthread -lgomp
srsRAN Project srsRAN Project is a complete ORAN-native 5G RAN solution created by Software Radio Systems (SRS). The srsRAN Project radio suite was formerly known as srsLTE and can be used for building your own software-defined radio (SDR) 4G/5G mobile network. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Mbps, More Is Better srsRAN Project 23.10.1-20240325 Test: PUSCH Processor Benchmark, Throughput Total Graviton4 96 Cores r8g.metal-24xl Graviton3 64 Cores r7g.metal Graviton2 64 Cores r6g.metal Graviton1 16 Cores a1.metal 400 800 1200 1600 2000 SE +/- 0.07, N = 3 SE +/- 0.03, N = 3 SE +/- 0.00, N = 3 SE +/- 0.03, N = 3 1961.7 1054.2 727.9 169.3 MIN: 1132.6 / MAX: 1961.8 MIN: 599.6 MIN: 416 1. (CXX) g++ options: -O3 -fno-trapping-math -fno-math-errno -ldl
LZ4 Compression OpenBenchmarking.org MB/s, More Is Better LZ4 Compression 1.10 Compression Level: 12 - Decompression Speed Graviton4 96 Cores r8g.metal-24xl Graviton3 64 Cores r7g.metal Graviton2 64 Cores r6g.metal Graviton1 16 Cores a1.metal 800 1600 2400 3200 4000 SE +/- 1.89, N = 3 SE +/- 0.17, N = 3 SE +/- 0.29, N = 3 SE +/- 2.61, N = 3 3595.5 3108.3 2207.6 1090.3 1. (CC) gcc options: -O3 -pthread
OpenBenchmarking.org MB/s, More Is Better LZ4 Compression 1.10 Compression Level: 12 - Compression Speed Graviton4 96 Cores r8g.metal-24xl Graviton3 64 Cores r7g.metal Graviton2 64 Cores r6g.metal Graviton1 16 Cores a1.metal 3 6 9 12 15 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 11.95 9.84 8.02 4.83 1. (CC) gcc options: -O3 -pthread
OpenBenchmarking.org MB/s, More Is Better LZ4 Compression 1.10 Compression Level: 9 - Decompression Speed Graviton4 96 Cores r8g.metal-24xl Graviton3 64 Cores r7g.metal Graviton2 64 Cores r6g.metal Graviton1 16 Cores a1.metal 800 1600 2400 3200 4000 SE +/- 0.40, N = 3 SE +/- 0.20, N = 3 SE +/- 0.31, N = 3 SE +/- 0.06, N = 3 3573.1 3088.6 2200.3 1102.0 1. (CC) gcc options: -O3 -pthread
OpenBenchmarking.org MB/s, More Is Better LZ4 Compression 1.10 Compression Level: 9 - Compression Speed Graviton4 96 Cores r8g.metal-24xl Graviton3 64 Cores r7g.metal Graviton2 64 Cores r6g.metal Graviton1 16 Cores a1.metal 9 18 27 36 45 SE +/- 0.01, N = 3 SE +/- 0.01, N = 3 SE +/- 0.03, N = 3 SE +/- 0.01, N = 3 39.80 33.23 25.50 13.72 1. (CC) gcc options: -O3 -pthread
PHPBench PHPBench is a benchmark suite for PHP. It performs a large number of simple tests in order to bench various aspects of the PHP interpreter. PHPBench can be used to compare hardware, operating systems, PHP versions, PHP accelerators and caches, compiler options, etc. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Score, More Is Better PHPBench 0.8.1 PHP Benchmark Suite Graviton4 96 Cores r8g.metal-24xl Graviton3 64 Cores r7g.metal Graviton2 64 Cores r6g.metal Graviton1 16 Cores a1.metal 200K 400K 600K 800K 1000K SE +/- 273.39, N = 3 SE +/- 964.81, N = 3 SE +/- 2661.11, N = 3 SE +/- 481.21, N = 3 829221 641106 444204 234211
PyPerformance OpenBenchmarking.org Milliseconds, Fewer Is Better PyPerformance 1.11 Benchmark: regex_compile Graviton4 96 Cores r8g.metal-24xl Graviton3 64 Cores r7g.metal Graviton2 64 Cores r6g.metal Graviton1 16 Cores a1.metal 100 200 300 400 500 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.33, N = 3 101 137 178 455
C-Ray OpenBenchmarking.org Seconds, Fewer Is Better C-Ray 2.0 Resolution: 1080p - Rays Per Pixel: 16 Graviton4 96 Cores r8g.metal-24xl Graviton3 64 Cores r7g.metal Graviton2 64 Cores r6g.metal Graviton1 16 Cores a1.metal 30 60 90 120 150 SE +/- 0.007, N = 3 SE +/- 0.010, N = 3 SE +/- 0.010, N = 3 SE +/- 0.248, N = 3 4.891 9.836 16.181 140.822 1. (CC) gcc options: -lpthread -lm
PyPerformance OpenBenchmarking.org Milliseconds, Fewer Is Better PyPerformance 1.11 Benchmark: pathlib Graviton4 96 Cores r8g.metal-24xl Graviton3 64 Cores r7g.metal Graviton2 64 Cores r6g.metal Graviton1 16 Cores a1.metal 20 40 60 80 100 SE +/- 0.00, N = 3 SE +/- 0.03, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 17.6 23.7 31.7 110.0
Coremark This is a test of EEMBC CoreMark processor benchmark. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Iterations/Sec, More Is Better Coremark 1.0 CoreMark Size 666 - Iterations Per Second Graviton4 96 Cores r8g.metal-24xl Graviton3 64 Cores r7g.metal Graviton2 64 Cores r6g.metal Graviton1 16 Cores a1.metal 600K 1200K 1800K 2400K 3000K SE +/- 64.82, N = 3 SE +/- 10334.27, N = 13 SE +/- 300.99, N = 3 SE +/- 24.05, N = 3 2732642.27 1535079.55 1176422.69 192273.03 1. (CC) gcc options: -O2 -lrt" -lrt
WebP Image Encode OpenBenchmarking.org MP/s, More Is Better WebP Image Encode 1.4 Encode Settings: Quality 100, Lossless Graviton4 96 Cores r8g.metal-24xl Graviton3 64 Cores r7g.metal Graviton2 64 Cores r6g.metal Graviton1 16 Cores a1.metal 0.2993 0.5986 0.8979 1.1972 1.4965 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 1.33 1.07 0.80 0.40 1. (CC) gcc options: -fvisibility=hidden -O2 -lm
PyPerformance OpenBenchmarking.org Milliseconds, Fewer Is Better PyPerformance 1.11 Benchmark: crypto_pyaes Graviton4 96 Cores r8g.metal-24xl Graviton3 64 Cores r7g.metal Graviton2 64 Cores r6g.metal Graviton1 16 Cores a1.metal 50 100 150 200 250 SE +/- 0.03, N = 3 SE +/- 0.03, N = 3 SE +/- 0.00, N = 3 SE +/- 0.33, N = 3 65.0 90.1 110.0 227.0
7-Zip Compression OpenBenchmarking.org MIPS, More Is Better 7-Zip Compression 24.05 Test: Decompression Rating Graviton4 96 Cores r8g.metal-24xl Graviton3 64 Cores r7g.metal Graviton2 64 Cores r6g.metal Graviton1 16 Cores a1.metal 110K 220K 330K 440K 550K SE +/- 158.70, N = 3 SE +/- 38.31, N = 3 SE +/- 32.36, N = 3 SE +/- 50.81, N = 3 494660 285729 233493 41244 1. (CXX) g++ options: -lpthread -ldl -O2 -fPIC
OpenBenchmarking.org MIPS, More Is Better 7-Zip Compression 24.05 Test: Compression Rating Graviton4 96 Cores r8g.metal-24xl Graviton3 64 Cores r7g.metal Graviton2 64 Cores r6g.metal Graviton1 16 Cores a1.metal 100K 200K 300K 400K 500K SE +/- 2385.17, N = 3 SE +/- 865.25, N = 3 SE +/- 159.81, N = 3 SE +/- 28.98, N = 3 479851 299377 233491 32697 1. (CXX) g++ options: -lpthread -ldl -O2 -fPIC
John The Ripper This is a benchmark of John The Ripper, which is a password cracker. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Real C/S, More Is Better John The Ripper 2023.03.14 Test: WPA PSK Graviton4 96 Cores r8g.metal-24xl Graviton3 64 Cores r7g.metal Graviton2 64 Cores r6g.metal Graviton1 16 Cores a1.metal 20K 40K 60K 80K 100K SE +/- 0.00, N = 3 SE +/- 6.00, N = 3 SE +/- 11.59, N = 3 SE +/- 1.33, N = 3 86281 52734 49975 7894 1. (CC) gcc options: -lssl -lcrypto -fopenmp -lm -lrt -lz -ldl -lcrypt
PyPerformance OpenBenchmarking.org Milliseconds, Fewer Is Better PyPerformance 1.11 Benchmark: json_loads Graviton4 96 Cores r8g.metal-24xl Graviton3 64 Cores r7g.metal Graviton2 64 Cores r6g.metal Graviton1 16 Cores a1.metal 16 32 48 64 80 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.10, N = 3 20.6 28.7 36.7 72.1
Stress-NG OpenBenchmarking.org Bogo Ops/s, More Is Better Stress-NG 0.17.08 Test: Power Math Graviton4 96 Cores r8g.metal-24xl Graviton3 64 Cores r7g.metal Graviton2 64 Cores r6g.metal Graviton1 16 Cores a1.metal 5K 10K 15K 20K 25K SE +/- 0.98, N = 3 SE +/- 0.10, N = 3 SE +/- 13.10, N = 3 SE +/- 0.08, N = 3 23418.40 13073.07 9242.02 1609.16 1. (CXX) g++ options: -O2 -std=gnu99 -lc -lm
OpenBenchmarking.org Bogo Ops/s, More Is Better Stress-NG 0.17.08 Test: Matrix 3D Math Graviton4 96 Cores r8g.metal-24xl Graviton3 64 Cores r7g.metal Graviton2 64 Cores r6g.metal Graviton1 16 Cores a1.metal 5K 10K 15K 20K 25K SE +/- 50.98, N = 3 SE +/- 6.42, N = 3 SE +/- 4.86, N = 3 SE +/- 3.67, N = 3 22886.62 9831.89 5845.88 838.94 1. (CXX) g++ options: -O2 -std=gnu99 -lc -lm
OpenBenchmarking.org Bogo Ops/s, More Is Better Stress-NG 0.17.08 Test: Vector Floating Point Graviton4 96 Cores r8g.metal-24xl Graviton3 64 Cores r7g.metal Graviton2 64 Cores r6g.metal Graviton1 16 Cores a1.metal 30K 60K 90K 120K 150K SE +/- 17.26, N = 3 SE +/- 29.81, N = 3 SE +/- 35.00, N = 3 SE +/- 0.38, N = 3 119727.57 76127.77 42695.46 3466.25 1. (CXX) g++ options: -O2 -std=gnu99 -lc -lm
OpenBenchmarking.org Bogo Ops/s, More Is Better Stress-NG 0.17.08 Test: Logarithmic Math Graviton4 96 Cores r8g.metal-24xl Graviton3 64 Cores r7g.metal Graviton2 64 Cores r6g.metal Graviton1 16 Cores a1.metal 11K 22K 33K 44K 55K SE +/- 1.22, N = 3 SE +/- 0.23, N = 3 SE +/- 0.91, N = 3 SE +/- 0.28, N = 3 50252.66 27799.23 18828.22 3302.68 1. (CXX) g++ options: -O2 -std=gnu99 -lc -lm
OpenBenchmarking.org Bogo Ops/s, More Is Better Stress-NG 0.17.08 Test: Trigonometric Math Graviton4 96 Cores r8g.metal-24xl Graviton3 64 Cores r7g.metal Graviton2 64 Cores r6g.metal Graviton1 16 Cores a1.metal 10K 20K 30K 40K 50K SE +/- 11.48, N = 3 SE +/- 1.81, N = 3 SE +/- 2.00, N = 3 SE +/- 0.04, N = 3 45810.37 26006.52 17338.17 2998.75 1. (CXX) g++ options: -O2 -std=gnu99 -lc -lm
OpenBenchmarking.org Bogo Ops/s, More Is Better Stress-NG 0.17.08 Test: Floating Point Graviton4 96 Cores r8g.metal-24xl Graviton3 64 Cores r7g.metal Graviton2 64 Cores r6g.metal Graviton1 16 Cores a1.metal 3K 6K 9K 12K 15K SE +/- 2.60, N = 3 SE +/- 0.20, N = 3 SE +/- 0.17, N = 3 SE +/- 1.60, N = 3 12455.02 7047.63 4700.48 852.80 1. (CXX) g++ options: -O2 -std=gnu99 -lc -lm
OpenBenchmarking.org Bogo Ops/s, More Is Better Stress-NG 0.17.08 Test: Memory Copying Graviton4 96 Cores r8g.metal-24xl Graviton3 64 Cores r7g.metal Graviton2 64 Cores r6g.metal Graviton1 16 Cores a1.metal 7K 14K 21K 28K 35K SE +/- 13.60, N = 3 SE +/- 1.92, N = 3 SE +/- 1.30, N = 3 SE +/- 0.09, N = 3 32525.12 20452.57 11890.83 1997.89 1. (CXX) g++ options: -O2 -std=gnu99 -lc -lm
OpenBenchmarking.org Bogo Ops/s, More Is Better Stress-NG 0.17.08 Test: Matrix Math Graviton4 96 Cores r8g.metal-24xl Graviton3 64 Cores r7g.metal Graviton2 64 Cores r6g.metal Graviton1 16 Cores a1.metal 140K 280K 420K 560K 700K SE +/- 9.55, N = 3 SE +/- 12.06, N = 3 SE +/- 7.14, N = 3 SE +/- 74.42, N = 3 631796.65 368563.70 284337.61 7343.99 1. (CXX) g++ options: -O2 -std=gnu99 -lc -lm
OpenBenchmarking.org Bogo Ops/s, More Is Better Stress-NG 0.17.08 Test: Vector Math Graviton4 96 Cores r8g.metal-24xl Graviton3 64 Cores r7g.metal Graviton2 64 Cores r6g.metal Graviton1 16 Cores a1.metal 90K 180K 270K 360K 450K SE +/- 85.60, N = 3 SE +/- 51.34, N = 3 SE +/- 6.90, N = 3 SE +/- 0.54, N = 3 438910.40 262952.32 165655.63 28508.49 1. (CXX) g++ options: -O2 -std=gnu99 -lc -lm
OpenBenchmarking.org Bogo Ops/s, More Is Better Stress-NG 0.17.08 Test: Fused Multiply-Add Graviton4 96 Cores r8g.metal-24xl Graviton3 64 Cores r7g.metal Graviton2 64 Cores r6g.metal Graviton1 16 Cores a1.metal 14M 28M 42M 56M 70M SE +/- 1063.93, N = 3 SE +/- 3002.12, N = 3 SE +/- 1364.97, N = 3 SE +/- 2472.21, N = 3 65585333.04 42419807.93 33894303.54 4928995.12 1. (CXX) g++ options: -O2 -std=gnu99 -lc -lm
miniFE MiniFE Finite Element is an application for unstructured implicit finite element codes. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org CG Mflops, More Is Better miniFE 2.2 Problem Size: Small Graviton4 96 Cores r8g.metal-24xl Graviton3 64 Cores r7g.metal Graviton2 64 Cores r6g.metal Graviton1 16 Cores a1.metal 15K 30K 45K 60K 75K SE +/- 233.77, N = 3 SE +/- 30.17, N = 3 SE +/- 11.77, N = 3 SE +/- 1.93, N = 3 70677.80 36601.50 22864.20 4159.12 1. (CXX) g++ options: -O3 -fopenmp -lmpi_cxx -lmpi
srsRAN Project srsRAN Project is a complete ORAN-native 5G RAN solution created by Software Radio Systems (SRS). The srsRAN Project radio suite was formerly known as srsLTE and can be used for building your own software-defined radio (SDR) 4G/5G mobile network. Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Mbps, More Is Better srsRAN Project 23.10.1-20240325 Test: PDSCH Processor Benchmark, Throughput Total Graviton4 96 Cores r8g.metal-24xl Graviton3 64 Cores r7g.metal Graviton2 64 Cores r6g.metal Graviton1 16 Cores a1.metal 5K 10K 15K 20K 25K SE +/- 15.31, N = 3 SE +/- 189.99, N = 3 SE +/- 88.68, N = 5 SE +/- 10.32, N = 3 22062.7 13395.4 8295.6 1307.3 1. (CXX) g++ options: -O3 -fno-trapping-math -fno-math-errno -ldl
OpenBenchmarking.org Mbps, More Is Better srsRAN Project 23.10.1-20240325 Test: PDSCH Processor Benchmark, Throughput Thread Graviton4 96 Cores r8g.metal-24xl Graviton3 64 Cores r7g.metal Graviton2 64 Cores r6g.metal Graviton1 16 Cores a1.metal 60 120 180 240 300 SE +/- 3.00, N = 3 SE +/- 0.15, N = 3 SE +/- 0.07, N = 3 SE +/- 0.21, N = 3 278.7 239.6 146.6 94.0 1. (CXX) g++ options: -O3 -fno-trapping-math -fno-math-errno -ldl
libavif avifenc This is a test of the AOMedia libavif library testing the encoding of a JPEG image to AV1 Image Format (AVIF). Learn more via the OpenBenchmarking.org test page.
OpenBenchmarking.org Seconds, Fewer Is Better libavif avifenc 1.0 Encoder Speed: 6, Lossless Graviton4 96 Cores r8g.metal-24xl Graviton3 64 Cores r7g.metal Graviton2 64 Cores r6g.metal Graviton1 16 Cores a1.metal 7 14 21 28 35 SE +/- 0.046, N = 3 SE +/- 0.086, N = 3 SE +/- 0.034, N = 3 SE +/- 0.098, N = 3 5.237 6.483 8.969 28.229 1. (CXX) g++ options: -O3 -fPIC -lm
WebP Image Encode OpenBenchmarking.org MP/s, More Is Better WebP Image Encode 1.4 Encode Settings: Quality 100, Highest Compression Graviton4 96 Cores r8g.metal-24xl Graviton3 64 Cores r7g.metal Graviton2 64 Cores r6g.metal Graviton1 16 Cores a1.metal 0.7245 1.449 2.1735 2.898 3.6225 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.00, N = 3 SE +/- 0.01, N = 3 3.22 2.64 2.07 1.43 1. (CC) gcc options: -fvisibility=hidden -O2 -lm
OpenBenchmarking.org MP/s, More Is Better WebP Image Encode 1.4 Encode Settings: Quality 100 Graviton4 96 Cores r8g.metal-24xl Graviton3 64 Cores r7g.metal Graviton2 64 Cores r6g.metal Graviton1 16 Cores a1.metal 2 4 6 8 10 SE +/- 0.01, N = 3 SE +/- 0.01, N = 3 SE +/- 0.01, N = 3 SE +/- 0.01, N = 3 7.82 6.72 5.15 3.25 1. (CC) gcc options: -fvisibility=hidden -O2 -lm
OpenBenchmarking.org MP/s, More Is Better WebP Image Encode 1.4 Encode Settings: Default Graviton4 96 Cores r8g.metal-24xl Graviton3 64 Cores r7g.metal Graviton2 64 Cores r6g.metal Graviton1 16 Cores a1.metal 3 6 9 12 15 SE +/- 0.02, N = 3 SE +/- 0.02, N = 3 SE +/- 0.00, N = 3 SE +/- 0.01, N = 3 11.61 9.87 7.55 4.63 1. (CC) gcc options: -fvisibility=hidden -O2 -lm
Graviton1 16 Cores a1.metal Processor: ARMv8 Cortex-A72 (16 Cores), Motherboard: Amazon EC2 a1.metal (1.0 BIOS), Memory: 32GB, Disk: 429GB Amazon Elastic Block Store, Network: Amazon Elastic
OS: Ubuntu 24.04, Kernel: 6.8.0-1009-aws (aarch64), Compiler: GCC 13.2.0, File-System: ext4
Kernel Notes: Transparent Huge Pages: madviseCompiler Notes: --build=aarch64-linux-gnu --disable-libquadmath --disable-libquadmath-support --disable-werror --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-fix-cortex-a53-843419 --enable-gnu-unique-object --enable-languages=c,ada,c++,go,d,fortran,objc,obj-c++,m2 --enable-libphobos-checking=release --enable-libstdcxx-backtrace --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-multiarch --enable-nls --enable-objc-gc=auto --enable-offload-defaulted --enable-offload-targets=nvptx-none=/build/gcc-13-dIwDw0/gcc-13-13.2.0/debian/tmp-nvptx/usr --enable-plugin --enable-shared --enable-threads=posix --host=aarch64-linux-gnu --program-prefix=aarch64-linux-gnu- --target=aarch64-linux-gnu --with-default-libstdcxx-abi=new --with-gcc-major-version-only --with-target-system-zlib=auto --without-cuda-driver -vJava Notes: OpenJDK Runtime Environment (build 11.0.23+9-post-Ubuntu-1ubuntu1)Python Notes: Python 3.12.3Security Notes: gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + reg_file_data_sampling: Not affected + retbleed: Not affected + spec_rstack_overflow: Not affected + spec_store_bypass: Not affected + spectre_v1: Mitigation of __user pointer sanitization + spectre_v2: Mitigation of Branch predictor hardening BHB + srbds: Not affected + tsx_async_abort: Not affected
Testing initiated at 22 July 2024 16:03 by user ubuntu.
Graviton2 64 Cores r6g.metal Processor: ARMv8 Neoverse-N1 (64 Cores), Motherboard: Amazon EC2 r6g.metal (1.0 BIOS), Memory: 8 x 64GB DDR4-3200MT/s, Disk: 429GB Amazon Elastic Block Store, Network: Amazon Elastic
OS: Ubuntu 24.04, Kernel: 6.8.0-1009-aws (aarch64), Compiler: GCC 13.2.0, File-System: ext4
Kernel Notes: Transparent Huge Pages: madviseCompiler Notes: --build=aarch64-linux-gnu --disable-libquadmath --disable-libquadmath-support --disable-werror --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-fix-cortex-a53-843419 --enable-gnu-unique-object --enable-languages=c,ada,c++,go,d,fortran,objc,obj-c++,m2 --enable-libphobos-checking=release --enable-libstdcxx-backtrace --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-multiarch --enable-nls --enable-objc-gc=auto --enable-offload-defaulted --enable-offload-targets=nvptx-none=/build/gcc-13-dIwDw0/gcc-13-13.2.0/debian/tmp-nvptx/usr --enable-plugin --enable-shared --enable-threads=posix --host=aarch64-linux-gnu --program-prefix=aarch64-linux-gnu- --target=aarch64-linux-gnu --with-default-libstdcxx-abi=new --with-gcc-major-version-only --with-target-system-zlib=auto --without-cuda-driver -vJava Notes: OpenJDK Runtime Environment (build 11.0.23+9-post-Ubuntu-1ubuntu1)Python Notes: Python 3.12.3Security Notes: gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + reg_file_data_sampling: Not affected + retbleed: Not affected + spec_rstack_overflow: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of __user pointer sanitization + spectre_v2: Mitigation of CSV2 BHB + srbds: Not affected + tsx_async_abort: Not affected
Testing initiated at 23 July 2024 20:03 by user ubuntu.
Graviton3 64 Cores r7g.metal Processor: ARMv8 Neoverse-V1 (64 Cores), Motherboard: Amazon EC2 r7g.metal (1.0 BIOS), Memory: 8 x 64GB DDR5-4400MT/s, Disk: 429GB Amazon Elastic Block Store, Network: Amazon Elastic
OS: Ubuntu 24.04, Kernel: 6.8.0-1009-aws (aarch64), Compiler: GCC 13.2.0, File-System: ext4
Kernel Notes: Transparent Huge Pages: madviseCompiler Notes: --build=aarch64-linux-gnu --disable-libquadmath --disable-libquadmath-support --disable-werror --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-fix-cortex-a53-843419 --enable-gnu-unique-object --enable-languages=c,ada,c++,go,d,fortran,objc,obj-c++,m2 --enable-libphobos-checking=release --enable-libstdcxx-backtrace --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-multiarch --enable-nls --enable-objc-gc=auto --enable-offload-defaulted --enable-offload-targets=nvptx-none=/build/gcc-13-dIwDw0/gcc-13-13.2.0/debian/tmp-nvptx/usr --enable-plugin --enable-shared --enable-threads=posix --host=aarch64-linux-gnu --program-prefix=aarch64-linux-gnu- --target=aarch64-linux-gnu --with-default-libstdcxx-abi=new --with-gcc-major-version-only --with-target-system-zlib=auto --without-cuda-driver -vJava Notes: OpenJDK Runtime Environment (build 11.0.23+9-post-Ubuntu-1ubuntu1)Python Notes: Python 3.12.3Security Notes: gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + reg_file_data_sampling: Not affected + retbleed: Not affected + spec_rstack_overflow: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of __user pointer sanitization + spectre_v2: Mitigation of CSV2 BHB + srbds: Not affected + tsx_async_abort: Not affected
Testing initiated at 24 July 2024 10:33 by user ubuntu.
Graviton4 96 Cores r8g.metal-24xl Processor: ARMv8 Neoverse-V2 (96 Cores), Motherboard: Amazon EC2 r8g.metal-24xl (1.0 BIOS), Memory: 12 x 64GB DDR5-5600MT/s, Disk: 429GB Amazon Elastic Block Store, Network: Amazon Elastic
OS: Ubuntu 24.04, Kernel: 6.8.0-1009-aws (aarch64), Compiler: GCC 13.2.0, File-System: ext4
Kernel Notes: Transparent Huge Pages: madviseCompiler Notes: --build=aarch64-linux-gnu --disable-libquadmath --disable-libquadmath-support --disable-werror --enable-checking=release --enable-clocale=gnu --enable-default-pie --enable-fix-cortex-a53-843419 --enable-gnu-unique-object --enable-languages=c,ada,c++,go,d,fortran,objc,obj-c++,m2 --enable-libphobos-checking=release --enable-libstdcxx-backtrace --enable-libstdcxx-debug --enable-libstdcxx-time=yes --enable-multiarch --enable-nls --enable-objc-gc=auto --enable-offload-defaulted --enable-offload-targets=nvptx-none=/build/gcc-13-dIwDw0/gcc-13-13.2.0/debian/tmp-nvptx/usr --enable-plugin --enable-shared --enable-threads=posix --host=aarch64-linux-gnu --program-prefix=aarch64-linux-gnu- --target=aarch64-linux-gnu --with-default-libstdcxx-abi=new --with-gcc-major-version-only --with-target-system-zlib=auto --without-cuda-driver -vJava Notes: OpenJDK Runtime Environment (build 11.0.23+9-post-Ubuntu-1ubuntu1)Python Notes: Python 3.12.3Security Notes: gather_data_sampling: Not affected + itlb_multihit: Not affected + l1tf: Not affected + mds: Not affected + meltdown: Not affected + mmio_stale_data: Not affected + reg_file_data_sampling: Not affected + retbleed: Not affected + spec_rstack_overflow: Not affected + spec_store_bypass: Mitigation of SSB disabled via prctl + spectre_v1: Mitigation of __user pointer sanitization + spectre_v2: Not affected + srbds: Not affected + tsx_async_abort: Not affected
Testing initiated at 23 July 2024 14:13 by user ubuntu.