x3d sun Benchmarks for a future article. AMD Ryzen 7 9800X3D 8-Core testing with a ASRock X870E Taichi (3.12.AS02 BIOS) and NVIDIA GeForce RTX 4070 12GB on Ubuntu 24.04 via the Phoronix Test Suite. a: Processor: AMD Ryzen 7 9800X3D 8-Core @ 6.22GHz (8 Cores / 16 Threads), Motherboard: ASRock X870E Taichi (3.12.AS02 BIOS), Chipset: AMD Device 14d8, Memory: 2 x 16GB DDR5-6000MT/s F5-6000J2836G16G, Disk: Western Digital WD_BLACK SN850X 2000GB, Graphics: NVIDIA GeForce RTX 4070 12GB, Audio: NVIDIA AD104 HD Audio, Monitor: DELL U2723QE, Network: Realtek Device 8126 + MEDIATEK Device 0717 OS: Ubuntu 24.04, Kernel: 6.8.0-49-generic (x86_64), Desktop: GNOME Shell 46.0, Display Server: X Server 1.21.1.11, Display Driver: NVIDIA 565.57.01, OpenGL: 4.6.0, OpenCL: OpenCL 3.0 CUDA 12.7.33, Compiler: GCC 13.2.0, File-System: ext4, Screen Resolution: 3840x2160 b: Processor: AMD Ryzen 7 9800X3D 8-Core @ 6.22GHz (8 Cores / 16 Threads), Motherboard: ASRock X870E Taichi (3.12.AS02 BIOS), Chipset: AMD Device 14d8, Memory: 2 x 16GB DDR5-6000MT/s F5-6000J2836G16G, Disk: Western Digital WD_BLACK SN850X 2000GB, Graphics: NVIDIA GeForce RTX 4070 12GB, Audio: NVIDIA AD104 HD Audio, Monitor: DELL U2723QE, Network: Realtek Device 8126 + MEDIATEK Device 0717 OS: Ubuntu 24.04, Kernel: 6.8.0-49-generic (x86_64), Desktop: GNOME Shell 46.0, Display Server: X Server 1.21.1.11, Display Driver: NVIDIA 565.57.01, OpenGL: 4.6.0, OpenCL: OpenCL 3.0 CUDA 12.7.33, Compiler: GCC 13.2.0, File-System: ext4, Screen Resolution: 3840x2160 Renaissance 0.16 Test: Scala Dotty ms < Lower Is Better a . 332.3 |=================================================================== b . 337.5 |==================================================================== Renaissance 0.16 Test: Random Forest ms < Lower Is Better a . 313.6 |=================================================================== b . 320.3 |==================================================================== Renaissance 0.16 Test: ALS Movie Lens ms < Lower Is Better a . 5675.8 |=================================================================== b . 5503.2 |================================================================= Renaissance 0.16 Test: Apache Spark ALS ms < Lower Is Better Renaissance 0.16 Test: Apache Spark Bayes ms < Lower Is Better a . 464.0 |==================================================================== b . 461.9 |==================================================================== Renaissance 0.16 Test: Savina Reactors.IO ms < Lower Is Better a . 3235.7 |================================================================== b . 3274.5 |=================================================================== Renaissance 0.16 Test: Apache Spark PageRank ms < Lower Is Better a . 1731.2 |================================================================== b . 1768.0 |=================================================================== Renaissance 0.16 Test: Finagle HTTP Requests ms < Lower Is Better a . 946.6 |============================================================= b . 1036.1 |=================================================================== Renaissance 0.16 Test: Gaussian Mixture Model ms < Lower Is Better a . 2754.2 |================================================================== b . 2785.8 |=================================================================== Renaissance 0.16 Test: In-Memory Database Shootout ms < Lower Is Better a . 1774.9 |=================================================================== b . 1719.0 |================================================================= Renaissance 0.16 Test: Akka Unbalanced Cobwebbed Tree ms < Lower Is Better a . 2997.6 |=================================================================== b . 2999.4 |=================================================================== Renaissance 0.16 Test: Genetic Algorithm Using Jenetics + Futures ms < Lower Is Better a . 629.9 |================================================================== b . 651.5 |==================================================================== Primesieve 12.6 Length: 1e12 Seconds < Lower Is Better a . 10.85 |==================================================================== b . 10.84 |==================================================================== Primesieve 12.6 Length: 1e13 Seconds < Lower Is Better a . 135.90 |=================================================================== b . 135.58 |=================================================================== Blender 4.3 Blend File: BMW27 - Compute: CPU-Only Seconds < Lower Is Better a . 84.28 |==================================================================== b . 84.85 |==================================================================== Blender 4.3 Blend File: Junkshop - Compute: CPU-Only Seconds < Lower Is Better a . 109.31 |================================================================== b . 110.24 |=================================================================== Blender 4.3 Blend File: BMW27 - Compute: NVIDIA OptiX Seconds < Lower Is Better a . 6.25 |==================================================================== b . 6.36 |===================================================================== Blender 4.3 Blend File: Classroom - Compute: CPU-Only Seconds < Lower Is Better a . 238.72 |================================================================== b . 240.88 |=================================================================== Blender 4.3 Blend File: Fishy Cat - Compute: CPU-Only Seconds < Lower Is Better a . 118.76 |=================================================================== b . 119.39 |=================================================================== Blender 4.3 Blend File: Barbershop - Compute: CPU-Only Seconds < Lower Is Better a . 822.15 |=================================================================== b . 824.33 |=================================================================== Blender 4.3 Blend File: Junkshop - Compute: NVIDIA OptiX Seconds < Lower Is Better a . 12.66 |==================================================================== b . 12.66 |==================================================================== Blender 4.3 Blend File: Classroom - Compute: NVIDIA OptiX Seconds < Lower Is Better a . 15.67 |==================================================================== b . 15.52 |=================================================================== Blender 4.3 Blend File: Fishy Cat - Compute: NVIDIA OptiX Seconds < Lower Is Better a . 11.51 |==================================================================== b . 11.55 |==================================================================== Blender 4.3 Blend File: Barbershop - Compute: NVIDIA OptiX Seconds < Lower Is Better a . 62.28 |==================================================================== b . 62.35 |==================================================================== Blender 4.3 Blend File: Pabellon Barcelona - Compute: CPU-Only Seconds < Lower Is Better a . 267.62 |=================================================================== b . 269.06 |=================================================================== Blender 4.3 Blend File: Pabellon Barcelona - Compute: NVIDIA OptiX Seconds < Lower Is Better a . 16.88 |==================================================================== b . 16.85 |==================================================================== OpenVINO 2024.5 Model: Face Detection FP16 - Device: CPU FPS > Higher Is Better a . 13.55 |==================================================================== b . 13.56 |==================================================================== OpenVINO 2024.5 Model: Face Detection FP16 - Device: CPU ms < Lower Is Better a . 295.13 |=================================================================== b . 294.80 |=================================================================== OpenVINO 2024.5 Model: Person Detection FP16 - Device: CPU FPS > Higher Is Better a . 125.14 |============================================================== b . 134.63 |=================================================================== OpenVINO 2024.5 Model: Person Detection FP16 - Device: CPU ms < Lower Is Better a . 31.93 |==================================================================== b . 29.67 |=============================================================== OpenVINO 2024.5 Model: Person Detection FP32 - Device: CPU FPS > Higher Is Better a . 125.85 |================================================================== b . 127.03 |=================================================================== OpenVINO 2024.5 Model: Person Detection FP32 - Device: CPU ms < Lower Is Better a . 31.75 |==================================================================== b . 31.45 |=================================================================== OpenVINO 2024.5 Model: Vehicle Detection FP16 - Device: CPU FPS > Higher Is Better a . 1074.93 |================================================================== b . 1053.24 |================================================================= OpenVINO 2024.5 Model: Vehicle Detection FP16 - Device: CPU ms < Lower Is Better a . 3.70 |==================================================================== b . 3.77 |===================================================================== OpenVINO 2024.5 Model: Face Detection FP16-INT8 - Device: CPU FPS > Higher Is Better a . 25.60 |==================================================================== b . 25.51 |==================================================================== OpenVINO 2024.5 Model: Face Detection FP16-INT8 - Device: CPU ms < Lower Is Better a . 156.19 |=================================================================== b . 156.63 |=================================================================== OpenVINO 2024.5 Model: Face Detection Retail FP16 - Device: CPU FPS > Higher Is Better a . 2720.37 |================================================================== b . 2649.11 |================================================================ OpenVINO 2024.5 Model: Face Detection Retail FP16 - Device: CPU ms < Lower Is Better a . 1.42 |==================================================================== b . 1.44 |===================================================================== OpenVINO 2024.5 Model: Road Segmentation ADAS FP16 - Device: CPU FPS > Higher Is Better a . 399.65 |=================================================================== b . 380.54 |================================================================ OpenVINO 2024.5 Model: Road Segmentation ADAS FP16 - Device: CPU ms < Lower Is Better a . 9.98 |================================================================= b . 10.48 |==================================================================== OpenVINO 2024.5 Model: Vehicle Detection FP16-INT8 - Device: CPU FPS > Higher Is Better a . 1432.67 |================================================================== b . 1429.01 |================================================================== OpenVINO 2024.5 Model: Vehicle Detection FP16-INT8 - Device: CPU ms < Lower Is Better a . 2.77 |===================================================================== b . 2.78 |===================================================================== OpenVINO 2024.5 Model: Weld Porosity Detection FP16 - Device: CPU FPS > Higher Is Better a . 1301.51 |================================================================== b . 1303.10 |================================================================== OpenVINO 2024.5 Model: Weld Porosity Detection FP16 - Device: CPU ms < Lower Is Better a . 6.04 |===================================================================== b . 6.04 |===================================================================== OpenVINO 2024.5 Model: Face Detection Retail FP16-INT8 - Device: CPU FPS > Higher Is Better a . 3828.93 |================================================================= b . 3868.70 |================================================================== OpenVINO 2024.5 Model: Face Detection Retail FP16-INT8 - Device: CPU ms < Lower Is Better a . 2.02 |===================================================================== b . 2.01 |===================================================================== OpenVINO 2024.5 Model: Road Segmentation ADAS FP16-INT8 - Device: CPU FPS > Higher Is Better a . 468.10 |=================================================================== b . 458.47 |================================================================== OpenVINO 2024.5 Model: Road Segmentation ADAS FP16-INT8 - Device: CPU ms < Lower Is Better a . 8.52 |==================================================================== b . 8.70 |===================================================================== OpenVINO 2024.5 Model: Machine Translation EN To DE FP16 - Device: CPU FPS > Higher Is Better a . 125.44 |=================================================================== b . 124.85 |=================================================================== OpenVINO 2024.5 Model: Machine Translation EN To DE FP16 - Device: CPU ms < Lower Is Better a . 31.85 |==================================================================== b . 32.00 |==================================================================== OpenVINO 2024.5 Model: Weld Porosity Detection FP16-INT8 - Device: CPU FPS > Higher Is Better a . 2441.60 |================================================================== b . 2427.32 |================================================================== OpenVINO 2024.5 Model: Weld Porosity Detection FP16-INT8 - Device: CPU ms < Lower Is Better a . 3.19 |===================================================================== b . 3.20 |===================================================================== OpenVINO 2024.5 Model: Person Vehicle Bike Detection FP16 - Device: CPU FPS > Higher Is Better a . 1275.14 |================================================================== b . 1273.96 |================================================================== OpenVINO 2024.5 Model: Person Vehicle Bike Detection FP16 - Device: CPU ms < Lower Is Better a . 3.09 |===================================================================== b . 3.10 |===================================================================== OpenVINO 2024.5 Model: Noise Suppression Poconet-Like FP16 - Device: CPU FPS > Higher Is Better a . 1228.77 |================================================================== b . 1220.46 |================================================================== OpenVINO 2024.5 Model: Noise Suppression Poconet-Like FP16 - Device: CPU ms < Lower Is Better a . 6.45 |===================================================================== b . 6.49 |===================================================================== OpenVINO 2024.5 Model: Handwritten English Recognition FP16 - Device: CPU FPS > Higher Is Better a . 559.03 |=================================================================== b . 550.55 |================================================================== OpenVINO 2024.5 Model: Handwritten English Recognition FP16 - Device: CPU ms < Lower Is Better a . 14.23 |=================================================================== b . 14.42 |==================================================================== OpenVINO 2024.5 Model: Person Re-Identification Retail FP16 - Device: CPU FPS > Higher Is Better a . 1612.55 |================================================================== b . 1608.60 |================================================================== OpenVINO 2024.5 Model: Person Re-Identification Retail FP16 - Device: CPU ms < Lower Is Better a . 2.44 |===================================================================== b . 2.44 |===================================================================== OpenVINO 2024.5 Model: Age Gender Recognition Retail 0013 FP16 - Device: CPU FPS > Higher Is Better a . 30648.41 |================================================================= b . 30751.32 |================================================================= OpenVINO 2024.5 Model: Age Gender Recognition Retail 0013 FP16 - Device: CPU ms < Lower Is Better a . 0.24 |===================================================================== b . 0.24 |===================================================================== OpenVINO 2024.5 Model: Handwritten English Recognition FP16-INT8 - Device: CPU FPS > Higher Is Better a . 565.22 |=================================================================== b . 564.93 |=================================================================== OpenVINO 2024.5 Model: Handwritten English Recognition FP16-INT8 - Device: CPU ms < Lower Is Better a . 14.05 |==================================================================== b . 14.06 |==================================================================== OpenVINO 2024.5 Model: Age Gender Recognition Retail 0013 FP16-INT8 - Device: CPU FPS > Higher Is Better a . 40630.00 |================================================================= b . 40555.42 |================================================================= OpenVINO 2024.5 Model: Age Gender Recognition Retail 0013 FP16-INT8 - Device: CPU ms < Lower Is Better a . 0.18 |===================================================================== b . 0.18 |===================================================================== Llama.cpp b4154 Backend: CPU BLAS - Model: Llama-3.1-Tulu-3-8B-Q8_0 - Test: Text Generation 128 Tokens Per Second > Higher Is Better a . 7.27 |===================================================================== b . 7.28 |===================================================================== Llama.cpp b4154 Backend: CPU BLAS - Model: Llama-3.1-Tulu-3-8B-Q8_0 - Test: Prompt Processing 512 Tokens Per Second > Higher Is Better a . 108.23 |================================================================== b . 109.30 |=================================================================== Llama.cpp b4154 Backend: CPU BLAS - Model: Llama-3.1-Tulu-3-8B-Q8_0 - Test: Prompt Processing 1024 Tokens Per Second > Higher Is Better a . 104.57 |=================================================================== b . 104.56 |=================================================================== Llama.cpp b4154 Backend: CPU BLAS - Model: Llama-3.1-Tulu-3-8B-Q8_0 - Test: Prompt Processing 2048 Tokens Per Second > Higher Is Better a . 96.53 |==================================================================== b . 96.74 |==================================================================== Llama.cpp b4154 Backend: CPU BLAS - Model: Mistral-7B-Instruct-v0.3-Q8_0 - Test: Text Generation 128 Tokens Per Second > Higher Is Better a . 7.68 |===================================================================== b . 7.67 |===================================================================== Llama.cpp b4154 Backend: CPU BLAS - Model: Mistral-7B-Instruct-v0.3-Q8_0 - Test: Prompt Processing 512 Tokens Per Second > Higher Is Better a . 108.99 |=================================================================== b . 108.38 |=================================================================== Llama.cpp b4154 Backend: CPU BLAS - Model: Mistral-7B-Instruct-v0.3-Q8_0 - Test: Prompt Processing 1024 Tokens Per Second > Higher Is Better a . 104.62 |=================================================================== b . 103.69 |================================================================== Llama.cpp b4154 Backend: CPU BLAS - Model: Mistral-7B-Instruct-v0.3-Q8_0 - Test: Prompt Processing 2048 Tokens Per Second > Higher Is Better a . 96.13 |==================================================================== b . 95.20 |=================================================================== Llama.cpp b4154 Backend: CPU BLAS - Model: granite-3.0-3b-a800m-instruct-Q8_0 - Test: Text Generation 128 Tokens Per Second > Higher Is Better a . 60.37 |==================================================================== b . 60.15 |==================================================================== Llama.cpp b4154 Backend: CPU BLAS - Model: granite-3.0-3b-a800m-instruct-Q8_0 - Test: Prompt Processing 512 Tokens Per Second > Higher Is Better a . 320.35 |=================================================================== b . 294.73 |============================================================== Llama.cpp b4154 Backend: CPU BLAS - Model: granite-3.0-3b-a800m-instruct-Q8_0 - Test: Prompt Processing 1024 Tokens Per Second > Higher Is Better a . 268.90 |=================================================================== b . 268.47 |=================================================================== Llama.cpp b4154 Backend: CPU BLAS - Model: granite-3.0-3b-a800m-instruct-Q8_0 - Test: Prompt Processing 2048 Tokens Per Second > Higher Is Better a . 242.14 |=============================================================== b . 255.78 |=================================================================== OpenVINO GenAI 2024.5 Model: Gemma-7b-int4-ov - Device: CPU tokens/s > Higher Is Better a . 10.76 |==================================================================== b . 10.79 |==================================================================== OpenVINO GenAI 2024.5 Model: Gemma-7b-int4-ov - Device: CPU - Time To First Token ms < Lower Is Better a . 107.55 |=================================================================== b . 105.09 |================================================================= OpenVINO GenAI 2024.5 Model: Gemma-7b-int4-ov - Device: CPU - Time Per Output Token ms < Lower Is Better a . 92.93 |==================================================================== b . 92.68 |==================================================================== OpenVINO GenAI 2024.5 Model: TinyLlama-1.1B-Chat-v1.0 - Device: CPU tokens/s > Higher Is Better a . 26.42 |==================================================================== b . 26.39 |==================================================================== OpenVINO GenAI 2024.5 Model: TinyLlama-1.1B-Chat-v1.0 - Device: CPU - Time To First Token ms < Lower Is Better a . 39.61 |==================================================================== b . 39.66 |==================================================================== OpenVINO GenAI 2024.5 Model: TinyLlama-1.1B-Chat-v1.0 - Device: CPU - Time Per Output Token ms < Lower Is Better a . 37.85 |==================================================================== b . 37.89 |==================================================================== OpenVINO GenAI 2024.5 Model: Falcon-7b-instruct-int4-ov - Device: CPU tokens/s > Higher Is Better a . 14.17 |==================================================================== b . 14.19 |==================================================================== OpenVINO GenAI 2024.5 Model: Falcon-7b-instruct-int4-ov - Device: CPU - Time To First Token ms < Lower Is Better a . 80.92 |==================================================================== b . 80.51 |==================================================================== OpenVINO GenAI 2024.5 Model: Falcon-7b-instruct-int4-ov - Device: CPU - Time Per Output Token ms < Lower Is Better a . 70.58 |==================================================================== b . 70.49 |==================================================================== OpenVINO GenAI 2024.5 Model: Phi-3-mini-128k-instruct-int4-ov - Device: CPU tokens/s > Higher Is Better a . 21.63 |==================================================================== b . 21.61 |==================================================================== OpenVINO GenAI 2024.5 Model: Phi-3-mini-128k-instruct-int4-ov - Device: CPU - Time To First Token ms < Lower Is Better a . 51.76 |==================================================================== b . 51.79 |==================================================================== OpenVINO GenAI 2024.5 Model: Phi-3-mini-128k-instruct-int4-ov - Device: CPU - Time Per Output Token ms < Lower Is Better a . 46.24 |==================================================================== b . 46.27 |====================================================================