AWS Graviton Benchmarks Benchmarks by Michael Larabel. 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 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 Stress-NG 0.17.08 Test: CPU Cache Bogo Ops/s > Higher Is Better Graviton1 16 Cores a1.metal ....... 631797.13 |==== Graviton4 96 Cores r8g.metal-24xl . 4735225.17 |=============================== Stress-NG 0.17.08 Test: Power Math Bogo Ops/s > Higher Is Better Graviton1 16 Cores a1.metal ....... 1609.16 |== Graviton4 96 Cores r8g.metal-24xl . 23418.40 |================================= Stress-NG 0.17.08 Test: Matrix Math Bogo Ops/s > Higher Is Better Graviton1 16 Cores a1.metal ....... 7343.99 | Graviton4 96 Cores r8g.metal-24xl . 631796.65 |================================ Stress-NG 0.17.08 Test: Vector Math Bogo Ops/s > Higher Is Better Graviton1 16 Cores a1.metal ....... 28508.49 |== Graviton4 96 Cores r8g.metal-24xl . 438910.40 |================================ Stress-NG 0.17.08 Test: Floating Point Bogo Ops/s > Higher Is Better Graviton1 16 Cores a1.metal ....... 852.80 |== Graviton4 96 Cores r8g.metal-24xl . 12455.02 |================================= Stress-NG 0.17.08 Test: Matrix 3D Math Bogo Ops/s > Higher Is Better Graviton1 16 Cores a1.metal ....... 838.94 |= Graviton4 96 Cores r8g.metal-24xl . 22886.62 |================================= Stress-NG 0.17.08 Test: Memory Copying Bogo Ops/s > Higher Is Better Graviton1 16 Cores a1.metal ....... 1997.89 |== Graviton4 96 Cores r8g.metal-24xl . 32525.12 |================================= Stress-NG 0.17.08 Test: Logarithmic Math Bogo Ops/s > Higher Is Better Graviton1 16 Cores a1.metal ....... 3302.68 |== Graviton4 96 Cores r8g.metal-24xl . 50252.66 |================================= Stress-NG 0.17.08 Test: Fused Multiply-Add Bogo Ops/s > Higher Is Better Graviton1 16 Cores a1.metal ....... 4928995.12 |== Graviton4 96 Cores r8g.metal-24xl . 65585333.04 |============================== Stress-NG 0.17.08 Test: Trigonometric Math Bogo Ops/s > Higher Is Better Graviton1 16 Cores a1.metal ....... 2998.75 |== Graviton4 96 Cores r8g.metal-24xl . 45810.37 |================================= Stress-NG 0.17.08 Test: Vector Floating Point Bogo Ops/s > Higher Is Better Graviton1 16 Cores a1.metal ....... 3466.25 |= Graviton4 96 Cores r8g.metal-24xl . 119727.57 |================================ OpenSSL Algorithm: SHA256 byte/s > Higher Is Better Graviton1 16 Cores a1.metal ....... 6151749933 |== Graviton4 96 Cores r8g.metal-24xl . 85604846770 |============================== OpenSSL Algorithm: SHA512 byte/s > Higher Is Better Graviton1 16 Cores a1.metal ....... 2505891877 |= Graviton4 96 Cores r8g.metal-24xl . 53203418283 |============================== OpenSSL Algorithm: ChaCha20 byte/s > Higher Is Better Graviton1 16 Cores a1.metal ....... 8285450333 |== Graviton4 96 Cores r8g.metal-24xl . 148791688760 |============================= OpenSSL Algorithm: AES-128-GCM byte/s > Higher Is Better Graviton1 16 Cores a1.metal ....... 30455741337 |== Graviton4 96 Cores r8g.metal-24xl . 444871148797 |============================= OpenSSL Algorithm: AES-256-GCM byte/s > Higher Is Better Graviton1 16 Cores a1.metal ....... 25773496457 |== Graviton4 96 Cores r8g.metal-24xl . 400130558450 |============================= OpenSSL Algorithm: ChaCha20-Poly1305 byte/s > Higher Is Better Graviton1 16 Cores a1.metal ....... 6554146770 |== Graviton4 96 Cores r8g.metal-24xl . 110699450003 |============================= miniFE 2.2 Problem Size: Small CG Mflops > Higher Is Better Graviton1 16 Cores a1.metal ....... 4159.12 |== Graviton4 96 Cores r8g.metal-24xl . 70677.80 |================================= High Performance Conjugate Gradient 3.1 X Y Z: 104 104 104 - RT: 60 GFLOP/s > Higher Is Better Graviton1 16 Cores a1.metal ....... 3.76867 |== Graviton4 96 Cores r8g.metal-24xl . 61.61950 |================================= ACES DGEMM 1.0 Sustained Floating-Point Rate GFLOP/s > Higher Is Better Graviton1 16 Cores a1.metal ....... 0.879250 | Graviton4 96 Cores r8g.metal-24xl . 60.635839 |================================ GraphicsMagick 1.3.43 Operation: Sharpen Iterations Per Minute > Higher Is Better Graviton1 16 Cores a1.metal ....... 26 |=== Graviton4 96 Cores r8g.metal-24xl . 325 |====================================== GraphicsMagick 1.3.43 Operation: Enhanced Iterations Per Minute > Higher Is Better Graviton1 16 Cores a1.metal ....... 20 |=== Graviton4 96 Cores r8g.metal-24xl . 294 |====================================== GraphicsMagick 1.3.43 Operation: Resizing Iterations Per Minute > Higher Is Better Graviton1 16 Cores a1.metal ....... 61 |======== Graviton4 96 Cores r8g.metal-24xl . 278 |====================================== GraphicsMagick 1.3.43 Operation: Noise-Gaussian Iterations Per Minute > Higher Is Better Graviton1 16 Cores a1.metal ....... 26 |===== Graviton4 96 Cores r8g.metal-24xl . 208 |====================================== GraphicsMagick 1.3.43 Operation: HWB Color Space Iterations Per Minute > Higher Is Better Graviton1 16 Cores a1.metal ....... 81 |========== Graviton4 96 Cores r8g.metal-24xl . 301 |====================================== Coremark 1.0 CoreMark Size 666 - Iterations Per Second Iterations/Sec > Higher Is Better Graviton1 16 Cores a1.metal ....... 192273.03 |== Graviton4 96 Cores r8g.metal-24xl . 2732642.27 |=============================== LZ4 Compression 1.10 Compression Level: 1 - Compression Speed MB/s > Higher Is Better Graviton1 16 Cores a1.metal ....... 237.68 |============ Graviton4 96 Cores r8g.metal-24xl . 697.89 |=================================== LZ4 Compression 1.10 Compression Level: 1 - Decompression Speed MB/s > Higher Is Better Graviton1 16 Cores a1.metal ....... 1514.7 |============= Graviton4 96 Cores r8g.metal-24xl . 3927.7 |=================================== LZ4 Compression 1.10 Compression Level: 2 - Compression Speed MB/s > Higher Is Better Graviton1 16 Cores a1.metal ....... 82.51 |========= Graviton4 96 Cores r8g.metal-24xl . 329.41 |=================================== LZ4 Compression 1.10 Compression Level: 2 - Decompression Speed MB/s > Higher Is Better Graviton1 16 Cores a1.metal ....... 1153.1 |============ Graviton4 96 Cores r8g.metal-24xl . 3326.1 |=================================== LZ4 Compression 1.10 Compression Level: 3 - Compression Speed MB/s > Higher Is Better Graviton1 16 Cores a1.metal ....... 37.30 |=========== Graviton4 96 Cores r8g.metal-24xl . 115.14 |=================================== LZ4 Compression 1.10 Compression Level: 3 - Decompression Speed MB/s > Higher Is Better Graviton1 16 Cores a1.metal ....... 1137.3 |=========== Graviton4 96 Cores r8g.metal-24xl . 3489.5 |=================================== LZ4 Compression 1.10 Compression Level: 9 - Compression Speed MB/s > Higher Is Better Graviton1 16 Cores a1.metal ....... 13.72 |============ Graviton4 96 Cores r8g.metal-24xl . 39.80 |==================================== LZ4 Compression 1.10 Compression Level: 9 - Decompression Speed MB/s > Higher Is Better Graviton1 16 Cores a1.metal ....... 1102.0 |=========== Graviton4 96 Cores r8g.metal-24xl . 3573.1 |=================================== LZ4 Compression 1.10 Compression Level: 12 - Compression Speed MB/s > Higher Is Better Graviton1 16 Cores a1.metal ....... 4.83 |=============== Graviton4 96 Cores r8g.metal-24xl . 11.95 |==================================== LZ4 Compression 1.10 Compression Level: 12 - Decompression Speed MB/s > Higher Is Better Graviton1 16 Cores a1.metal ....... 1090.3 |=========== Graviton4 96 Cores r8g.metal-24xl . 3595.5 |=================================== srsRAN Project 23.10.1-20240325 Test: PDSCH Processor Benchmark, Throughput Total Mbps > Higher Is Better Graviton1 16 Cores a1.metal ....... 1307.3 |== Graviton4 96 Cores r8g.metal-24xl . 22062.7 |================================== srsRAN Project 23.10.1-20240325 Test: PUSCH Processor Benchmark, Throughput Total Mbps > Higher Is Better Graviton1 16 Cores a1.metal ....... 169.3 |=== Graviton4 96 Cores r8g.metal-24xl . 1961.7 |=================================== srsRAN Project 23.10.1-20240325 Test: PDSCH Processor Benchmark, Throughput Thread Mbps > Higher Is Better Graviton1 16 Cores a1.metal ....... 94.0 |============ Graviton4 96 Cores r8g.metal-24xl . 278.7 |==================================== srsRAN Project 23.10.1-20240325 Test: PUSCH Processor Benchmark, Throughput Thread Mbps > Higher Is Better Graviton1 16 Cores a1.metal ....... 18.7 |========== Graviton4 96 Cores r8g.metal-24xl . 69.4 |===================================== 7-Zip Compression 24.05 Test: Compression Rating MIPS > Higher Is Better Graviton1 16 Cores a1.metal ....... 32697 |== Graviton4 96 Cores r8g.metal-24xl . 479851 |=================================== 7-Zip Compression 24.05 Test: Decompression Rating MIPS > Higher Is Better Graviton1 16 Cores a1.metal ....... 41244 |=== Graviton4 96 Cores r8g.metal-24xl . 494660 |=================================== WebP Image Encode 1.4 Encode Settings: Default MP/s > Higher Is Better Graviton1 16 Cores a1.metal ....... 4.63 |============== Graviton4 96 Cores r8g.metal-24xl . 11.61 |==================================== WebP Image Encode 1.4 Encode Settings: Quality 100 MP/s > Higher Is Better Graviton1 16 Cores a1.metal ....... 3.25 |=============== Graviton4 96 Cores r8g.metal-24xl . 7.82 |===================================== WebP Image Encode 1.4 Encode Settings: Quality 100, Lossless MP/s > Higher Is Better Graviton1 16 Cores a1.metal ....... 0.40 |=========== Graviton4 96 Cores r8g.metal-24xl . 1.33 |===================================== WebP Image Encode 1.4 Encode Settings: Quality 100, Highest Compression MP/s > Higher Is Better Graviton1 16 Cores a1.metal ....... 1.43 |================ Graviton4 96 Cores r8g.metal-24xl . 3.22 |===================================== Stockfish 16.1 Chess Benchmark Nodes Per Second > Higher Is Better Graviton1 16 Cores a1.metal ....... 2493057 |= Graviton4 96 Cores r8g.metal-24xl . 118337918 |================================ GROMACS 2024 Implementation: MPI CPU - Input: water_GMX50_bare Ns Per Day > Higher Is Better Graviton1 16 Cores a1.metal ....... 0.313 |== Graviton4 96 Cores r8g.metal-24xl . 6.926 |==================================== Apache Cassandra 4.1.3 Test: Writes Op/s > Higher Is Better Graviton1 16 Cores a1.metal ....... 26348 |==== Graviton4 96 Cores r8g.metal-24xl . 239658 |=================================== RocksDB 9.0 Test: Random Read Op/s > Higher Is Better Graviton1 16 Cores a1.metal ....... 23252821 |= Graviton4 96 Cores r8g.metal-24xl . 527458547 |================================ RocksDB 9.0 Test: Update Random Op/s > Higher Is Better Graviton1 16 Cores a1.metal ....... 164616 |==== Graviton4 96 Cores r8g.metal-24xl . 1327423 |================================== RocksDB 9.0 Test: Read While Writing Op/s > Higher Is Better Graviton1 16 Cores a1.metal ....... 1268650 |==== Graviton4 96 Cores r8g.metal-24xl . 11560819 |================================= RocksDB 9.0 Test: Read Random Write Random Op/s > Higher Is Better Graviton1 16 Cores a1.metal ....... 574458 |=== Graviton4 96 Cores r8g.metal-24xl . 7047101 |================================== Speedb 2.7 Test: Random Read Op/s > Higher Is Better Graviton1 16 Cores a1.metal ....... 23286630 |= Graviton4 96 Cores r8g.metal-24xl . 640938471 |================================ Speedb 2.7 Test: Update Random Op/s > Higher Is Better Graviton1 16 Cores a1.metal ....... 128167 |====== Graviton4 96 Cores r8g.metal-24xl . 729398 |=================================== Speedb 2.7 Test: Read While Writing Op/s > Higher Is Better Graviton1 16 Cores a1.metal ....... 2433641 |======= Graviton4 96 Cores r8g.metal-24xl . 11490911 |================================= Speedb 2.7 Test: Read Random Write Random Op/s > Higher Is Better Graviton1 16 Cores a1.metal ....... 552621 |==== Graviton4 96 Cores r8g.metal-24xl . 5342911 |================================== John The Ripper 2023.03.14 Test: bcrypt Real C/S > Higher Is Better Graviton1 16 Cores a1.metal ....... 7842 |=== Graviton4 96 Cores r8g.metal-24xl . 85647 |==================================== John The Ripper 2023.03.14 Test: WPA PSK Real C/S > Higher Is Better Graviton1 16 Cores a1.metal ....... 7894 |=== Graviton4 96 Cores r8g.metal-24xl . 86281 |==================================== John The Ripper 2023.03.14 Test: Blowfish Real C/S > Higher Is Better Graviton1 16 Cores a1.metal ....... 8003 |=== Graviton4 96 Cores r8g.metal-24xl . 85667 |==================================== John The Ripper 2023.03.14 Test: HMAC-SHA512 Real C/S > Higher Is Better Graviton1 16 Cores a1.metal ....... 10321667 |== Graviton4 96 Cores r8g.metal-24xl . 158238333 |================================ John The Ripper 2023.03.14 Test: MD5 Real C/S > Higher Is Better Graviton1 16 Cores a1.metal ....... 190426 |=== Graviton4 96 Cores r8g.metal-24xl . 2367333 |================================== Numpy Benchmark Score > Higher Is Better Graviton1 16 Cores a1.metal ....... 104.59 |======= Graviton4 96 Cores r8g.metal-24xl . 495.78 |=================================== PHPBench 0.8.1 PHP Benchmark Suite Score > Higher Is Better Graviton1 16 Cores a1.metal ....... 234211 |========== Graviton4 96 Cores r8g.metal-24xl . 829221 |=================================== PostgreSQL 16 Scaling Factor: 100 - Clients: 1000 - Mode: Read Only TPS > Higher Is Better Graviton1 16 Cores a1.metal ....... 101870 |= Graviton4 96 Cores r8g.metal-24xl . 2893264 |================================== PostgreSQL 16 Scaling Factor: 100 - Clients: 1000 - Mode: Read Write TPS > Higher Is Better Graviton1 16 Cores a1.metal ....... 4632 |===================================== Graviton4 96 Cores r8g.metal-24xl . 4303 |================================== PyPerformance 1.11 Benchmark: pathlib Milliseconds < Lower Is Better Graviton1 16 Cores a1.metal ....... 110.0 |==================================== Graviton4 96 Cores r8g.metal-24xl . 17.6 |====== PyPerformance 1.11 Benchmark: xml_etree Milliseconds < Lower Is Better Graviton1 16 Cores a1.metal ....... 230.0 |==================================== Graviton4 96 Cores r8g.metal-24xl . 58.2 |========= PyPerformance 1.11 Benchmark: json_loads Milliseconds < Lower Is Better Graviton1 16 Cores a1.metal ....... 72.1 |===================================== Graviton4 96 Cores r8g.metal-24xl . 20.6 |=========== PyPerformance 1.11 Benchmark: crypto_pyaes Milliseconds < Lower Is Better Graviton1 16 Cores a1.metal ....... 227.0 |==================================== Graviton4 96 Cores r8g.metal-24xl . 65.0 |========== PyPerformance 1.11 Benchmark: regex_compile Milliseconds < Lower Is Better Graviton1 16 Cores a1.metal ....... 455 |====================================== Graviton4 96 Cores r8g.metal-24xl . 101 |======== PyPerformance 1.11 Benchmark: python_startup Milliseconds < Lower Is Better Graviton1 16 Cores a1.metal ....... 25.90 |==================================== Graviton4 96 Cores r8g.metal-24xl . 7.55 |========== PostgreSQL 16 Scaling Factor: 100 - Clients: 1000 - Mode: Read Only - Average Latency ms < Lower Is Better Graviton1 16 Cores a1.metal ....... 9.816 |==================================== Graviton4 96 Cores r8g.metal-24xl . 0.346 |= PostgreSQL 16 Scaling Factor: 100 - Clients: 1000 - Mode: Read Write - Average Latency ms < Lower Is Better Graviton1 16 Cores a1.metal ....... 215.89 |================================= Graviton4 96 Cores r8g.metal-24xl . 232.43 |=================================== libavif avifenc 1.0 Encoder Speed: 0 Seconds < Lower Is Better Graviton1 16 Cores a1.metal ....... 668.50 |=================================== Graviton4 96 Cores r8g.metal-24xl . 130.20 |======= libavif avifenc 1.0 Encoder Speed: 2 Seconds < Lower Is Better Graviton1 16 Cores a1.metal ....... 381.11 |=================================== Graviton4 96 Cores r8g.metal-24xl . 80.24 |======= libavif avifenc 1.0 Encoder Speed: 6 Seconds < Lower Is Better Graviton1 16 Cores a1.metal ....... 21.288 |=================================== Graviton4 96 Cores r8g.metal-24xl . 2.777 |===== libavif avifenc 1.0 Encoder Speed: 6, Lossless Seconds < Lower Is Better Graviton1 16 Cores a1.metal ....... 28.229 |=================================== Graviton4 96 Cores r8g.metal-24xl . 5.237 |====== Timed Godot Game Engine Compilation 4.0 Time To Compile Seconds < Lower Is Better Graviton1 16 Cores a1.metal ....... 1674.47 |================================== Graviton4 96 Cores r8g.metal-24xl . 126.35 |=== Timed Node.js Compilation 21.7.2 Time To Compile Seconds < Lower Is Better Graviton1 16 Cores a1.metal ....... 4663.28 |================================== Graviton4 96 Cores r8g.metal-24xl . 289.27 |== Timed PHP Compilation 8.3.4 Time To Compile Seconds < Lower Is Better Graviton1 16 Cores a1.metal ....... 479.64 |=================================== Graviton4 96 Cores r8g.metal-24xl . 90.00 |======= C-Ray 2.0 Resolution: 4K - Rays Per Pixel: 16 Seconds < Lower Is Better Graviton1 16 Cores a1.metal ....... 566.95 |=================================== Graviton4 96 Cores r8g.metal-24xl . 18.93 |= C-Ray 2.0 Resolution: 5K - Rays Per Pixel: 16 Seconds < Lower Is Better Graviton1 16 Cores a1.metal ....... 1008.69 |================================== Graviton4 96 Cores r8g.metal-24xl . 33.68 |= C-Ray 2.0 Resolution: 1080p - Rays Per Pixel: 16 Seconds < Lower Is Better Graviton1 16 Cores a1.metal ....... 140.822 |================================== Graviton4 96 Cores r8g.metal-24xl . 4.891 |= Blender 4.0.2 Blend File: BMW27 - Compute: CPU-Only Seconds < Lower Is Better Graviton1 16 Cores a1.metal ....... 599.39 |=================================== Graviton4 96 Cores r8g.metal-24xl . 33.72 |== Blender 4.0.2 Blend File: Classroom - Compute: CPU-Only Seconds < Lower Is Better Graviton1 16 Cores a1.metal ....... 1289.61 |================================== Graviton4 96 Cores r8g.metal-24xl . 69.40 |== Blender 4.0.2 Blend File: Fishy Cat - Compute: CPU-Only Seconds < Lower Is Better Graviton1 16 Cores a1.metal ....... 952.43 |=================================== Graviton4 96 Cores r8g.metal-24xl . 64.07 |==