jan tr AMD Ryzen Threadripper 7980X 64-Cores testing with a ASUS Pro WS TRX50-SAGE WIFI (0217 BIOS) and AMD Radeon RX 7900 XT 20GB on Ubuntu 23.10 via the Phoronix Test Suite. a: Processor: AMD Ryzen Threadripper 7980X 64-Cores @ 8.21GHz (64 Cores / 128 Threads), Motherboard: ASUS Pro WS TRX50-SAGE WIFI (0217 BIOS), Chipset: AMD Device 14a4, Memory: 128GB, Disk: 2000GB Corsair MP700 PRO + 1000GB Western Digital WDS100T1X0E-00AFY0, Graphics: AMD Radeon RX 7900 XT 20GB (2025/1249MHz), Audio: Realtek ALC1220, Monitor: DELL U2723QE, Network: Aquantia Device 04c0 + Intel I226-LM + MEDIATEK MT7922 802.11ax PCI OS: Ubuntu 23.10, Kernel: 6.7.0-060700rc2daily20231126-generic (x86_64), Desktop: GNOME Shell 45.0, Display Server: X Server 1.21.1.7 + Wayland, OpenGL: 4.6 Mesa 23.2.1-1ubuntu3 (LLVM 15.0.7 DRM 3.56), Compiler: GCC 13.2.0, File-System: ext4, Screen Resolution: 3840x2160 b: Processor: AMD Ryzen Threadripper 7980X 64-Cores @ 8.21GHz (64 Cores / 128 Threads), Motherboard: ASUS Pro WS TRX50-SAGE WIFI (0217 BIOS), Chipset: AMD Device 14a4, Memory: 128GB, Disk: 2000GB Corsair MP700 PRO + 1000GB Western Digital WDS100T1X0E-00AFY0, Graphics: AMD Radeon RX 7900 XT 20GB (2025/1249MHz), Audio: Realtek ALC1220, Monitor: DELL U2723QE, Network: Aquantia Device 04c0 + Intel I226-LM + MEDIATEK MT7922 802.11ax PCI OS: Ubuntu 23.10, Kernel: 6.7.0-060700rc2daily20231126-generic (x86_64), Desktop: GNOME Shell 45.0, Display Server: X Server 1.21.1.7 + Wayland, OpenGL: 4.6 Mesa 23.2.1-1ubuntu3 (LLVM 15.0.7 DRM 3.56), Compiler: GCC 13.2.0, File-System: ext4, Screen Resolution: 3840x2160 c: Processor: AMD Ryzen Threadripper 7980X 64-Cores @ 8.21GHz (64 Cores / 128 Threads), Motherboard: ASUS Pro WS TRX50-SAGE WIFI (0217 BIOS), Chipset: AMD Device 14a4, Memory: 128GB, Disk: 2000GB Corsair MP700 PRO + 1000GB Western Digital WDS100T1X0E-00AFY0, Graphics: AMD Radeon RX 7900 XT 20GB (2025/1249MHz), Audio: Realtek ALC1220, Monitor: DELL U2723QE, Network: Aquantia Device 04c0 + Intel I226-LM + MEDIATEK MT7922 802.11ax PCI OS: Ubuntu 23.10, Kernel: 6.7.0-060700rc2daily20231126-generic (x86_64), Desktop: GNOME Shell 45.0, Display Server: X Server 1.21.1.7 + Wayland, OpenGL: 4.6 Mesa 23.2.1-1ubuntu3 (LLVM 15.0.7 DRM 3.56), Compiler: GCC 13.2.0, File-System: ext4, Screen Resolution: 3840x2160 d: Processor: AMD Ryzen Threadripper 7980X 64-Cores @ 8.21GHz (64 Cores / 128 Threads), Motherboard: ASUS Pro WS TRX50-SAGE WIFI (0217 BIOS), Chipset: AMD Device 14a4, Memory: 128GB, Disk: 2000GB Corsair MP700 PRO + 1000GB Western Digital WDS100T1X0E-00AFY0, Graphics: AMD Radeon RX 7900 XT 20GB (2025/1249MHz), Audio: Realtek ALC1220, Monitor: DELL U2723QE, Network: Aquantia Device 04c0 + Intel I226-LM + MEDIATEK MT7922 802.11ax PCI OS: Ubuntu 23.10, Kernel: 6.7.0-060700rc2daily20231126-generic (x86_64), Desktop: GNOME Shell 45.0, Display Server: X Server 1.21.1.7 + Wayland, OpenGL: 4.6 Mesa 23.2.1-1ubuntu3 (LLVM 15.0.7 DRM 3.56), Compiler: GCC 13.2.0, File-System: ext4, Screen Resolution: 3840x2160 LeelaChessZero 0.30 Backend: BLAS Nodes Per Second > Higher Is Better a . 515 |==================================================================== b . 505 |=================================================================== c . 530 |====================================================================== d . 513 |==================================================================== LeelaChessZero 0.30 Backend: Eigen Nodes Per Second > Higher Is Better a . 453 |================================================================= b . 489 |====================================================================== c . 450 |================================================================ d . 453 |================================================================= Neural Magic DeepSparse 1.6 Model: NLP Document Classification, oBERT base uncased on IMDB - Scenario: Asynchronous Multi-Stream items/sec > Higher Is Better a . 21.89 |==================================================================== b . 21.93 |==================================================================== c . 21.91 |==================================================================== d . 21.89 |==================================================================== Neural Magic DeepSparse 1.6 Model: NLP Document Classification, oBERT base uncased on IMDB - Scenario: Asynchronous Multi-Stream ms/batch < Lower Is Better a . 365.46 |=================================================================== b . 364.81 |=================================================================== c . 365.05 |=================================================================== d . 364.35 |=================================================================== Neural Magic DeepSparse 1.6 Model: NLP Text Classification, BERT base uncased SST2, Sparse INT8 - Scenario: Asynchronous Multi-Stream items/sec > Higher Is Better a . 915.99 |=================================================================== b . 912.78 |=================================================================== c . 912.54 |=================================================================== d . 912.78 |=================================================================== Neural Magic DeepSparse 1.6 Model: NLP Text Classification, BERT base uncased SST2, Sparse INT8 - Scenario: Asynchronous Multi-Stream ms/batch < Lower Is Better a . 8.7254 |=================================================================== b . 8.7560 |=================================================================== c . 8.7584 |=================================================================== d . 8.7558 |=================================================================== Neural Magic DeepSparse 1.6 Model: ResNet-50, Baseline - Scenario: Asynchronous Multi-Stream items/sec > Higher Is Better a . 274.18 |=================================================================== b . 274.78 |=================================================================== c . 274.62 |=================================================================== d . 274.73 |=================================================================== Neural Magic DeepSparse 1.6 Model: ResNet-50, Baseline - Scenario: Asynchronous Multi-Stream ms/batch < Lower Is Better a . 29.16 |==================================================================== b . 29.10 |==================================================================== c . 29.11 |==================================================================== d . 29.11 |==================================================================== Neural Magic DeepSparse 1.6 Model: ResNet-50, Sparse INT8 - Scenario: Asynchronous Multi-Stream items/sec > Higher Is Better a . 2264.81 |================================================================= b . 2262.13 |================================================================= c . 2304.98 |================================================================== d . 2278.92 |================================================================= Neural Magic DeepSparse 1.6 Model: ResNet-50, Sparse INT8 - Scenario: Asynchronous Multi-Stream ms/batch < Lower Is Better a . 3.5237 |=================================================================== b . 3.5287 |=================================================================== c . 3.4631 |================================================================== d . 3.5027 |=================================================================== Neural Magic DeepSparse 1.6 Model: CV Detection, YOLOv5s COCO - Scenario: Asynchronous Multi-Stream items/sec > Higher Is Better a . 127.29 |=================================================================== b . 128.04 |=================================================================== c . 128.14 |=================================================================== d . 127.01 |================================================================== Neural Magic DeepSparse 1.6 Model: CV Detection, YOLOv5s COCO - Scenario: Asynchronous Multi-Stream ms/batch < Lower Is Better a . 62.83 |==================================================================== b . 62.43 |=================================================================== c . 62.40 |=================================================================== d . 62.97 |==================================================================== Neural Magic DeepSparse 1.6 Model: BERT-Large, NLP Question Answering - Scenario: Asynchronous Multi-Stream items/sec > Higher Is Better a . 27.36 |==================================================================== b . 27.30 |==================================================================== c . 27.33 |==================================================================== d . 27.31 |==================================================================== Neural Magic DeepSparse 1.6 Model: BERT-Large, NLP Question Answering - Scenario: Asynchronous Multi-Stream ms/batch < Lower Is Better a . 292.08 |=================================================================== b . 292.22 |=================================================================== c . 292.21 |=================================================================== d . 292.09 |=================================================================== Neural Magic DeepSparse 1.6 Model: CV Classification, ResNet-50 ImageNet - Scenario: Asynchronous Multi-Stream items/sec > Higher Is Better a . 274.88 |=================================================================== b . 274.68 |=================================================================== c . 274.71 |=================================================================== d . 274.49 |=================================================================== Neural Magic DeepSparse 1.6 Model: CV Classification, ResNet-50 ImageNet - Scenario: Asynchronous Multi-Stream ms/batch < Lower Is Better a . 29.09 |==================================================================== b . 29.11 |==================================================================== c . 29.11 |==================================================================== d . 29.13 |==================================================================== Neural Magic DeepSparse 1.6 Model: CV Detection, YOLOv5s COCO, Sparse INT8 - Scenario: Asynchronous Multi-Stream items/sec > Higher Is Better a . 131.12 |=================================================================== b . 131.39 |=================================================================== c . 131.08 |=================================================================== d . 130.30 |================================================================== Neural Magic DeepSparse 1.6 Model: CV Detection, YOLOv5s COCO, Sparse INT8 - Scenario: Asynchronous Multi-Stream ms/batch < Lower Is Better a . 61.00 |==================================================================== b . 60.86 |=================================================================== c . 60.97 |==================================================================== d . 61.38 |==================================================================== Neural Magic DeepSparse 1.6 Model: NLP Text Classification, DistilBERT mnli - Scenario: Asynchronous Multi-Stream items/sec > Higher Is Better a . 187.91 |=================================================================== b . 188.17 |=================================================================== c . 187.92 |=================================================================== d . 187.74 |=================================================================== Neural Magic DeepSparse 1.6 Model: NLP Text Classification, DistilBERT mnli - Scenario: Asynchronous Multi-Stream ms/batch < Lower Is Better a . 42.54 |==================================================================== b . 42.48 |==================================================================== c . 42.56 |==================================================================== d . 42.60 |==================================================================== Neural Magic DeepSparse 1.6 Model: CV Segmentation, 90% Pruned YOLACT Pruned - Scenario: Asynchronous Multi-Stream items/sec > Higher Is Better a . 40.17 |==================================================================== b . 40.19 |==================================================================== c . 40.24 |==================================================================== d . 40.18 |==================================================================== Neural Magic DeepSparse 1.6 Model: CV Segmentation, 90% Pruned YOLACT Pruned - Scenario: Asynchronous Multi-Stream ms/batch < Lower Is Better a . 198.92 |=================================================================== b . 198.92 |=================================================================== c . 198.54 |=================================================================== d . 199.00 |=================================================================== Neural Magic DeepSparse 1.6 Model: BERT-Large, NLP Question Answering, Sparse INT8 - Scenario: Asynchronous Multi-Stream items/sec > Higher Is Better a . 401.68 |=================================================================== b . 402.97 |=================================================================== c . 402.02 |=================================================================== d . 402.18 |=================================================================== Neural Magic DeepSparse 1.6 Model: BERT-Large, NLP Question Answering, Sparse INT8 - Scenario: Asynchronous Multi-Stream ms/batch < Lower Is Better a . 19.90 |==================================================================== b . 19.84 |==================================================================== c . 19.89 |==================================================================== d . 19.88 |==================================================================== Neural Magic DeepSparse 1.6 Model: NLP Token Classification, BERT base uncased conll2003 - Scenario: Asynchronous Multi-Stream items/sec > Higher Is Better a . 22.10 |==================================================================== b . 22.25 |==================================================================== c . 22.22 |==================================================================== d . 22.00 |=================================================================== Neural Magic DeepSparse 1.6 Model: NLP Token Classification, BERT base uncased conll2003 - Scenario: Asynchronous Multi-Stream ms/batch < Lower Is Better a . 361.90 |=================================================================== b . 359.34 |================================================================== c . 359.99 |================================================================== d . 363.63 |=================================================================== rav1e 0.7 Speed: 1 Frames Per Second > Higher Is Better a . 1.382 |==================================================================== b . 1.382 |==================================================================== c . 1.376 |==================================================================== d . 1.384 |==================================================================== rav1e 0.7 Speed: 5 Frames Per Second > Higher Is Better a . 5.760 |==================================================================== b . 5.747 |==================================================================== c . 5.748 |==================================================================== d . 5.768 |==================================================================== rav1e 0.7 Speed: 6 Frames Per Second > Higher Is Better a . 7.713 |==================================================================== b . 7.750 |==================================================================== c . 7.690 |=================================================================== d . 7.729 |==================================================================== rav1e 0.7 Speed: 10 Frames Per Second > Higher Is Better a . 18.43 |==================================================================== b . 18.22 |=================================================================== c . 18.35 |==================================================================== d . 18.40 |==================================================================== SVT-AV1 1.8 Encoder Mode: Preset 4 - Input: Bosphorus 4K Frames Per Second > Higher Is Better a . 9.915 |=================================================================== b . 9.876 |=================================================================== c . 9.950 |==================================================================== d . 9.998 |==================================================================== SVT-AV1 1.8 Encoder Mode: Preset 8 - Input: Bosphorus 4K Frames Per Second > Higher Is Better a . 96.14 |=================================================================== b . 95.98 |=================================================================== c . 97.29 |==================================================================== d . 95.98 |=================================================================== SVT-AV1 1.8 Encoder Mode: Preset 12 - Input: Bosphorus 4K Frames Per Second > Higher Is Better a . 226.38 |================================================================== b . 226.71 |================================================================== c . 217.36 |=============================================================== d . 229.79 |=================================================================== SVT-AV1 1.8 Encoder Mode: Preset 13 - Input: Bosphorus 4K Frames Per Second > Higher Is Better a . 233.40 |================================================================== b . 235.40 |=================================================================== c . 230.91 |================================================================== d . 224.79 |================================================================ SVT-AV1 1.8 Encoder Mode: Preset 4 - Input: Bosphorus 1080p Frames Per Second > Higher Is Better a . 26.90 |==================================================================== b . 26.75 |=================================================================== c . 26.90 |==================================================================== d . 27.05 |==================================================================== SVT-AV1 1.8 Encoder Mode: Preset 8 - Input: Bosphorus 1080p Frames Per Second > Higher Is Better a . 190.00 |================================================================== b . 188.00 |================================================================== c . 188.50 |================================================================== d . 191.46 |=================================================================== SVT-AV1 1.8 Encoder Mode: Preset 12 - Input: Bosphorus 1080p Frames Per Second > Higher Is Better a . 679.13 |================================================================== b . 686.95 |=================================================================== c . 684.73 |=================================================================== d . 660.56 |================================================================ SVT-AV1 1.8 Encoder Mode: Preset 13 - Input: Bosphorus 1080p Frames Per Second > Higher Is Better a . 798.11 |================================================================== b . 810.67 |=================================================================== c . 766.92 |=============================================================== d . 772.84 |================================================================ Xmrig 6.21 Variant: KawPow - Hash Count: 1M H/s > Higher Is Better a . 43081.2 |================================================================== b . 43084.9 |================================================================== c . 42260.1 |================================================================ d . 43368.9 |================================================================== Xmrig 6.21 Variant: Monero - Hash Count: 1M H/s > Higher Is Better a . 42620.3 |================================================================ b . 44163.8 |================================================================== c . 43269.4 |================================================================= d . 43097.9 |================================================================ Xmrig 6.21 Variant: Wownero - Hash Count: 1M H/s > Higher Is Better a . 51015.2 |================================================================== b . 47129.8 |============================================================= c . 49907.7 |================================================================= d . 49729.0 |================================================================ Xmrig 6.21 Variant: GhostRider - Hash Count: 1M H/s > Higher Is Better a . 11640.2 |=============================================================== b . 12124.6 |================================================================== c . 11899.9 |================================================================= d . 11952.1 |================================================================= Xmrig 6.21 Variant: CryptoNight-Heavy - Hash Count: 1M H/s > Higher Is Better a . 43183.5 |================================================================= b . 44058.7 |================================================================== c . 43573.0 |================================================================= d . 43740.7 |================================================================== Xmrig 6.21 Variant: CryptoNight-Femto UPX2 - Hash Count: 1M H/s > Higher Is Better a . 42573.1 |============================================================== b . 45093.8 |================================================================== c . 44012.1 |================================================================ d . 43979.2 |================================================================