Intel Xeon Gold 6421N testing with a Quanta Cloud S6Q-MB-MPS (3A10.uh BIOS) and ASPEED on Ubuntu 22.04 via the Phoronix Test Suite.
Compare your own system(s) to this result file with the
Phoronix Test Suite by running the command:
phoronix-test-suite benchmark 2307311-NE-NEWXEON6232
new xeon
Intel Xeon Gold 6421N testing with a Quanta Cloud S6Q-MB-MPS (3A10.uh BIOS) and ASPEED on Ubuntu 22.04 via the Phoronix Test Suite.
a:
Processor: Intel Xeon Gold 6421N @ 3.60GHz (32 Cores / 64 Threads), Motherboard: Quanta Cloud S6Q-MB-MPS (3A10.uh BIOS), Chipset: Intel Device 1bce, Memory: 512GB, Disk: 3 x 3841GB Micron_9300_MTFDHAL3T8TDP, Graphics: ASPEED, Monitor: VGA HDMI, Network: 4 x Intel E810-C for QSFP
OS: Ubuntu 22.04, Kernel: 5.15.0-47-generic (x86_64), Desktop: GNOME Shell 42.4, Display Server: X Server 1.21.1.3, Vulkan: 1.2.204, Compiler: GCC 11.2.0, File-System: ext4, Screen Resolution: 1600x1200
b:
Processor: Intel Xeon Gold 6421N @ 3.60GHz (32 Cores / 64 Threads), Motherboard: Quanta Cloud S6Q-MB-MPS (3A10.uh BIOS), Chipset: Intel Device 1bce, Memory: 512GB, Disk: 3 x 3841GB Micron_9300_MTFDHAL3T8TDP, Graphics: ASPEED, Monitor: VGA HDMI, Network: 4 x Intel E810-C for QSFP
OS: Ubuntu 22.04, Kernel: 5.15.0-47-generic (x86_64), Desktop: GNOME Shell 42.4, Display Server: X Server 1.21.1.3, Vulkan: 1.2.204, Compiler: GCC 11.2.0, File-System: ext4, Screen Resolution: 1600x1200
Apache IoTDB 1.1.2
Device Count: 100 - Batch Size Per Write: 100 - Sensor Count: 200
Average Latency > Higher Is Better
a . 31.83 |=================================================
b . 43.86 |====================================================================
Apache IoTDB 1.1.2
Device Count: 100 - Batch Size Per Write: 100 - Sensor Count: 200
point/sec > Higher Is Better
a . 43074031.84 |==============================================================
b . 34191814.86 |=================================================
Stress-NG 0.15.10
Test: CPU Cache
Bogo Ops/s > Higher Is Better
a . 1537111.20 |===================================================
b . 1885833.11 |===============================================================
libxsmm 2-1.17-3645
M N K: 256
GFLOPS/s > Higher Is Better
a . 879.6 |====================================================================
b . 758.9 |===========================================================
Apache IoTDB 1.1.2
Device Count: 200 - Batch Size Per Write: 100 - Sensor Count: 200
Average Latency > Higher Is Better
a . 29.54 |================================================================
b . 31.63 |====================================================================
Apache IoTDB 1.1.2
Device Count: 100 - Batch Size Per Write: 100 - Sensor Count: 500
Average Latency > Higher Is Better
a . 69.08 |================================================================
b . 73.56 |====================================================================
Apache IoTDB 1.1.2
Device Count: 500 - Batch Size Per Write: 1 - Sensor Count: 500
Average Latency > Higher Is Better
a . 22.97 |====================================================================
b . 21.63 |================================================================
HeFFTe - Highly Efficient FFT for Exascale 2.3
Test: c2c - Backend: Stock - Precision: double - X Y Z: 128
GFLOP/s > Higher Is Better
a . 46.64 |================================================================
b . 49.52 |====================================================================
Neural Magic DeepSparse 1.5
Model: BERT-Large, NLP Question Answering - Scenario: Asynchronous Multi-Stream
items/sec > Higher Is Better
a . 35.15 |================================================================
b . 37.33 |====================================================================
Redis 7.0.12 + memtier_benchmark 2.0
Protocol: Redis - Clients: 100 - Set To Get Ratio: 1:10
Ops/sec > Higher Is Better
a . 2447092.01 |===============================================================
b . 2304730.19 |===========================================================
Apache IoTDB 1.1.2
Device Count: 200 - Batch Size Per Write: 100 - Sensor Count: 200
point/sec > Higher Is Better
a . 54224351.10 |==============================================================
b . 51199962.11 |===========================================================
Neural Magic DeepSparse 1.5
Model: BERT-Large, NLP Question Answering - Scenario: Asynchronous Multi-Stream
ms/batch < Lower Is Better
a . 453.48 |===================================================================
b . 428.67 |===============================================================
Apache IoTDB 1.1.2
Device Count: 100 - Batch Size Per Write: 100 - Sensor Count: 500
point/sec > Higher Is Better
a . 59041436.64 |==============================================================
b . 56018457.87 |===========================================================
Apache IoTDB 1.1.2
Device Count: 500 - Batch Size Per Write: 1 - Sensor Count: 500
point/sec > Higher Is Better
a . 1916642.90 |============================================================
b . 2009050.46 |===============================================================
Stress-NG 0.15.10
Test: Cloning
Bogo Ops/s > Higher Is Better
a . 9740.57 |==================================================================
b . 9326.09 |===============================================================
Neural Magic DeepSparse 1.5
Model: NLP Text Classification, BERT base uncased SST2 - Scenario: Asynchronous Multi-Stream
items/sec > Higher Is Better
a . 137.38 |================================================================
b . 143.44 |===================================================================
Neural Magic DeepSparse 1.5
Model: NLP Text Classification, BERT base uncased SST2 - Scenario: Asynchronous Multi-Stream
ms/batch < Lower Is Better
a . 116.38 |===================================================================
b . 111.50 |================================================================
Apache IoTDB 1.1.2
Device Count: 500 - Batch Size Per Write: 1 - Sensor Count: 200
Average Latency > Higher Is Better
a . 9.49 |==================================================================
b . 9.87 |=====================================================================
HeFFTe - Highly Efficient FFT for Exascale 2.3
Test: r2c - Backend: Stock - Precision: float - X Y Z: 256
GFLOP/s > Higher Is Better
a . 157.87 |================================================================
b . 164.05 |===================================================================
Apache IoTDB 1.1.2
Device Count: 500 - Batch Size Per Write: 1 - Sensor Count: 200
point/sec > Higher Is Better
a . 1576432.25 |===============================================================
b . 1521587.40 |=============================================================
HeFFTe - Highly Efficient FFT for Exascale 2.3
Test: c2c - Backend: FFTW - Precision: double - X Y Z: 128
GFLOP/s > Higher Is Better
a . 64.43 |====================================================================
b . 62.30 |==================================================================
Stress-NG 0.15.10
Test: Futex
Bogo Ops/s > Higher Is Better
a . 1541676.36 |===============================================================
b . 1492979.46 |=============================================================
srsRAN Project 23.5
Test: PUSCH Processor Benchmark, Throughput Total
Mbps > Higher Is Better
a . 5372.9 |=================================================================
b . 5543.7 |===================================================================
Stress-NG 0.15.10
Test: Pipe
Bogo Ops/s > Higher Is Better
a . 35837711.85 |============================================================
b . 36852791.12 |==============================================================
HeFFTe - Highly Efficient FFT for Exascale 2.3
Test: r2c - Backend: FFTW - Precision: float - X Y Z: 256
GFLOP/s > Higher Is Better
a . 149.83 |=================================================================
b . 154.05 |===================================================================
Apache IoTDB 1.1.2
Device Count: 100 - Batch Size Per Write: 1 - Sensor Count: 200
Average Latency > Higher Is Better
a . 14.58 |==================================================================
b . 14.98 |====================================================================
Apache IoTDB 1.1.2
Device Count: 200 - Batch Size Per Write: 1 - Sensor Count: 200
Average Latency > Higher Is Better
a . 11.86 |==================================================================
b . 12.18 |====================================================================
Stress-NG 0.15.10
Test: SENDFILE
Bogo Ops/s > Higher Is Better
a . 582724.63 |==============================================================
b . 598173.56 |================================================================
Redis 7.0.12 + memtier_benchmark 2.0
Protocol: Redis - Clients: 100 - Set To Get Ratio: 1:5
Ops/sec > Higher Is Better
a . 2285996.17 |===============================================================
b . 2227152.02 |=============================================================
Stress-NG 0.15.10
Test: Matrix Math
Bogo Ops/s > Higher Is Better
a . 160653.44 |================================================================
b . 156668.43 |==============================================================
Apache IoTDB 1.1.2
Device Count: 500 - Batch Size Per Write: 100 - Sensor Count: 500
point/sec > Higher Is Better
a . 67607191.64 |==============================================================
b . 65935725.67 |============================================================
Apache IoTDB 1.1.2
Device Count: 200 - Batch Size Per Write: 100 - Sensor Count: 500
Average Latency > Higher Is Better
a . 101.25 |===================================================================
b . 98.87 |=================================================================
Apache IoTDB 1.1.2
Device Count: 200 - Batch Size Per Write: 1 - Sensor Count: 500
point/sec > Higher Is Better
a . 1505080.34 |===============================================================
b . 1469808.89 |==============================================================
Apache IoTDB 1.1.2
Device Count: 200 - Batch Size Per Write: 100 - Sensor Count: 500
point/sec > Higher Is Better
a . 45677447.24 |=============================================================
b . 46726912.46 |==============================================================
Liquid-DSP 1.6
Threads: 16 - Buffer Length: 256 - Filter Length: 512
samples/s > Higher Is Better
a . 243940000 |===============================================================
b . 248820000 |================================================================
Apache IoTDB 1.1.2
Device Count: 100 - Batch Size Per Write: 1 - Sensor Count: 200
point/sec > Higher Is Better
a . 710382.44 |================================================================
b . 697217.55 |===============================================================
Neural Magic DeepSparse 1.5
Model: NLP Token Classification, BERT base uncased conll2003 - Scenario: Asynchronous Multi-Stream
items/sec > Higher Is Better
a . 33.94 |===================================================================
b . 34.54 |====================================================================
Neural Magic DeepSparse 1.5
Model: NLP Token Classification, BERT base uncased conll2003 - Scenario: Asynchronous Multi-Stream
ms/batch < Lower Is Better
a . 468.80 |===================================================================
b . 460.67 |==================================================================
srsRAN Project 23.5
Test: PUSCH Processor Benchmark, Throughput Thread
Mbps > Higher Is Better
a . 240.4 |====================================================================
b . 236.3 |===================================================================
Stress-NG 0.15.10
Test: IO_uring
Bogo Ops/s > Higher Is Better
a . 1529665.98 |===============================================================
b . 1503623.79 |==============================================================
Liquid-DSP 1.6
Threads: 16 - Buffer Length: 256 - Filter Length: 57
samples/s > Higher Is Better
a . 848435000 |===============================================================
b . 862195000 |================================================================
HeFFTe - Highly Efficient FFT for Exascale 2.3
Test: r2c - Backend: Stock - Precision: double - X Y Z: 128
GFLOP/s > Higher Is Better
a . 92.40 |====================================================================
b . 90.99 |===================================================================
Neural Magic DeepSparse 1.5
Model: NLP Text Classification, DistilBERT mnli - Scenario: Asynchronous Multi-Stream
items/sec > Higher Is Better
a . 295.83 |==================================================================
b . 299.93 |===================================================================
Neural Magic DeepSparse 1.5
Model: NLP Text Classification, DistilBERT mnli - Scenario: Asynchronous Multi-Stream
ms/batch < Lower Is Better
a . 54.07 |====================================================================
b . 53.33 |===================================================================
Apache IoTDB 1.1.2
Device Count: 500 - Batch Size Per Write: 100 - Sensor Count: 200
point/sec > Higher Is Better
a . 56894390.61 |==============================================================
b . 56137174.70 |=============================================================
Stress-NG 0.15.10
Test: Socket Activity
Bogo Ops/s > Higher Is Better
a . 24947.14 |================================================================
b . 25282.31 |=================================================================
Apache IoTDB 1.1.2
Device Count: 200 - Batch Size Per Write: 1 - Sensor Count: 500
Average Latency > Higher Is Better
a . 26.29 |===================================================================
b . 26.64 |====================================================================
Liquid-DSP 1.6
Threads: 32 - Buffer Length: 256 - Filter Length: 512
samples/s > Higher Is Better
a . 383555000 |================================================================
b . 378650000 |===============================================================
VVenC 1.9
Video Input: Bosphorus 4K - Video Preset: Fast
Frames Per Second > Higher Is Better
a . 5.842 |===================================================================
b . 5.917 |====================================================================
Timed LLVM Compilation 16.0
Build System: Unix Makefiles
Seconds < Lower Is Better
a . 323.86 |===================================================================
b . 319.85 |==================================================================
HeFFTe - Highly Efficient FFT for Exascale 2.3
Test: r2c - Backend: Stock - Precision: float - X Y Z: 128
GFLOP/s > Higher Is Better
a . 149.94 |==================================================================
b . 151.80 |===================================================================
Neural Magic DeepSparse 1.5
Model: CV Detection, YOLOv5s COCO, Sparse INT8 - Scenario: Asynchronous Multi-Stream
items/sec > Higher Is Better
a . 208.85 |==================================================================
b . 211.23 |===================================================================
Neural Magic DeepSparse 1.5
Model: CV Detection, YOLOv5s COCO, Sparse INT8 - Scenario: Asynchronous Multi-Stream
ms/batch < Lower Is Better
a . 76.58 |====================================================================
b . 75.72 |===================================================================
libxsmm 2-1.17-3645
M N K: 128
GFLOPS/s > Higher Is Better
a . 1211.8 |==================================================================
b . 1225.0 |===================================================================
HeFFTe - Highly Efficient FFT for Exascale 2.3
Test: c2c - Backend: FFTW - Precision: double - X Y Z: 256
GFLOP/s > Higher Is Better
a . 38.93 |====================================================================
b . 38.52 |===================================================================
libxsmm 2-1.17-3645
M N K: 32
GFLOPS/s > Higher Is Better
a . 440.0 |===================================================================
b . 444.6 |====================================================================
Redis 7.0.12 + memtier_benchmark 2.0
Protocol: Redis - Clients: 50 - Set To Get Ratio: 1:10
Ops/sec > Higher Is Better
a . 2316281.26 |===============================================================
b . 2293467.62 |==============================================================
HeFFTe - Highly Efficient FFT for Exascale 2.3
Test: c2c - Backend: FFTW - Precision: float - X Y Z: 256
GFLOP/s > Higher Is Better
a . 76.03 |====================================================================
b . 75.30 |===================================================================
VVenC 1.9
Video Input: Bosphorus 1080p - Video Preset: Fast
Frames Per Second > Higher Is Better
a . 16.10 |===================================================================
b . 16.25 |====================================================================
Stress-NG 0.15.10
Test: Atomic
Bogo Ops/s > Higher Is Better
a . 133.83 |===================================================================
b . 132.61 |==================================================================
Stress-NG 0.15.10
Test: Semaphores
Bogo Ops/s > Higher Is Better
a . 62126446.21 |==============================================================
b . 61651485.43 |==============================================================
HeFFTe - Highly Efficient FFT for Exascale 2.3
Test: c2c - Backend: Stock - Precision: double - X Y Z: 256
GFLOP/s > Higher Is Better
a . 38.96 |====================================================================
b . 38.68 |====================================================================
libxsmm 2-1.17-3645
M N K: 64
GFLOPS/s > Higher Is Better
a . 833.8 |====================================================================
b . 839.9 |====================================================================
srsRAN Project 23.5
Test: Downlink Processor Benchmark
Mbps > Higher Is Better
a . 705.8 |====================================================================
b . 710.9 |====================================================================
Apache IoTDB 1.1.2
Device Count: 100 - Batch Size Per Write: 1 - Sensor Count: 500
Average Latency > Higher Is Better
a . 28.27 |====================================================================
b . 28.45 |====================================================================
Stress-NG 0.15.10
Test: MMAP
Bogo Ops/s > Higher Is Better
a . 861.28 |===================================================================
b . 856.14 |===================================================================
HeFFTe - Highly Efficient FFT for Exascale 2.3
Test: r2c - Backend: FFTW - Precision: double - X Y Z: 128
GFLOP/s > Higher Is Better
a . 121.79 |===================================================================
b . 122.46 |===================================================================
Palabos 2.3
Grid Size: 400
Mega Site Updates Per Second > Higher Is Better
a . 287.27 |===================================================================
b . 285.76 |===================================================================
Apache IoTDB 1.1.2
Device Count: 100 - Batch Size Per Write: 1 - Sensor Count: 500
point/sec > Higher Is Better
a . 1191500.88 |===============================================================
b . 1185338.02 |===============================================================
HeFFTe - Highly Efficient FFT for Exascale 2.3
Test: c2c - Backend: FFTW - Precision: float - X Y Z: 128
GFLOP/s > Higher Is Better
a . 131.66 |===================================================================
b . 130.98 |===================================================================
HeFFTe - Highly Efficient FFT for Exascale 2.3
Test: r2c - Backend: FFTW - Precision: float - X Y Z: 128
GFLOP/s > Higher Is Better
a . 207.24 |===================================================================
b . 206.22 |===================================================================
Laghos 3.1
Test: Triple Point Problem
Major Kernels Total Rate > Higher Is Better
a . 177.78 |===================================================================
b . 176.92 |===================================================================
Apache IoTDB 1.1.2
Device Count: 500 - Batch Size Per Write: 100 - Sensor Count: 500
Average Latency > Higher Is Better
a . 68.34 |====================================================================
b . 68.01 |====================================================================
Neural Magic DeepSparse 1.5
Model: CV Segmentation, 90% Pruned YOLACT Pruned - Scenario: Asynchronous Multi-Stream
ms/batch < Lower Is Better
a . 345.15 |===================================================================
b . 343.52 |===================================================================
Neural Magic DeepSparse 1.5
Model: CV Segmentation, 90% Pruned YOLACT Pruned - Scenario: Asynchronous Multi-Stream
items/sec > Higher Is Better
a . 46.33 |====================================================================
b . 46.55 |====================================================================
Stress-NG 0.15.10
Test: Fused Multiply-Add
Bogo Ops/s > Higher Is Better
a . 34197705.63 |==============================================================
b . 34050669.23 |==============================================================
High Performance Conjugate Gradient 3.1
X Y Z: 160 160 160 - RT: 60
GFLOP/s > Higher Is Better
a . 27.51 |====================================================================
b . 27.40 |====================================================================
Stress-NG 0.15.10
Test: Function Call
Bogo Ops/s > Higher Is Better
a . 22028.03 |=================================================================
b . 22106.49 |=================================================================
Apache IoTDB 1.1.2
Device Count: 500 - Batch Size Per Write: 100 - Sensor Count: 200
Average Latency > Higher Is Better
a . 31.58 |====================================================================
b . 31.69 |====================================================================
HeFFTe - Highly Efficient FFT for Exascale 2.3
Test: r2c - Backend: FFTW - Precision: double - X Y Z: 512
GFLOP/s > Higher Is Better
a . 74.47 |====================================================================
b . 74.71 |====================================================================
Liquid-DSP 1.6
Threads: 32 - Buffer Length: 256 - Filter Length: 57
samples/s > Higher Is Better
a . 1328100000 |===============================================================
b . 1323900000 |===============================================================
Stress-NG 0.15.10
Test: NUMA
Bogo Ops/s > Higher Is Better
a . 390.87 |===================================================================
b . 392.08 |===================================================================
Stress-NG 0.15.10
Test: Mutex
Bogo Ops/s > Higher Is Better
a . 15147444.51 |==============================================================
b . 15192892.59 |==============================================================
HeFFTe - Highly Efficient FFT for Exascale 2.3
Test: c2c - Backend: Stock - Precision: float - X Y Z: 128
GFLOP/s > Higher Is Better
a . 85.74 |====================================================================
b . 85.49 |====================================================================
Neural Magic DeepSparse 1.5
Model: NLP Question Answering, BERT base uncased SQuaD 12layer Pruned90 - Scenario: Asynchronous Multi-Stream
ms/batch < Lower Is Better
a . 131.45 |===================================================================
b . 131.07 |===================================================================
Neural Magic DeepSparse 1.5
Model: NLP Question Answering, BERT base uncased SQuaD 12layer Pruned90 - Scenario: Asynchronous Multi-Stream
items/sec > Higher Is Better
a . 121.69 |===================================================================
b . 122.04 |===================================================================
Stress-NG 0.15.10
Test: Wide Vector Math
Bogo Ops/s > Higher Is Better
a . 1745029.27 |===============================================================
b . 1750003.43 |===============================================================
Apache IoTDB 1.1.2
Device Count: 200 - Batch Size Per Write: 1 - Sensor Count: 200
point/sec > Higher Is Better
a . 1045806.81 |===============================================================
b . 1042859.03 |===============================================================
Liquid-DSP 1.6
Threads: 64 - Buffer Length: 256 - Filter Length: 57
samples/s > Higher Is Better
a . 1728850000 |===============================================================
b . 1733700000 |===============================================================
Neural Magic DeepSparse 1.5
Model: NLP Sentiment Analysis, 80% Pruned Quantized BERT Base Uncased - Scenario: Asynchronous Multi-Stream
items/sec > Higher Is Better
a . 390.91 |===================================================================
b . 391.91 |===================================================================
Neural Magic DeepSparse 1.5
Model: NLP Sentiment Analysis, 80% Pruned Quantized BERT Base Uncased - Scenario: Asynchronous Multi-Stream
ms/batch < Lower Is Better
a . 40.91 |====================================================================
b . 40.81 |====================================================================
VVenC 1.9
Video Input: Bosphorus 4K - Video Preset: Faster
Frames Per Second > Higher Is Better
a . 11.02 |====================================================================
b . 10.99 |====================================================================
Redis 7.0.12 + memtier_benchmark 2.0
Protocol: Redis - Clients: 50 - Set To Get Ratio: 1:5
Ops/sec > Higher Is Better
a . 2211638.65 |===============================================================
b . 2217192.12 |===============================================================
Timed GDB GNU Debugger Compilation 10.2
Time To Compile
Seconds < Lower Is Better
a . 41.91 |====================================================================
b . 42.01 |====================================================================
Stress-NG 0.15.10
Test: Glibc C String Functions
Bogo Ops/s > Higher Is Better
a . 26067360.60 |==============================================================
b . 26125214.84 |==============================================================
High Performance Conjugate Gradient 3.1
X Y Z: 104 104 104 - RT: 60
GFLOP/s > Higher Is Better
a . 27.78 |====================================================================
b . 27.84 |====================================================================
HeFFTe - Highly Efficient FFT for Exascale 2.3
Test: c2c - Backend: Stock - Precision: float - X Y Z: 256
GFLOP/s > Higher Is Better
a . 75.09 |====================================================================
b . 74.93 |====================================================================
OpenFOAM 10
Input: drivaerFastback, Small Mesh Size - Execution Time
Seconds < Lower Is Better
a . 67.71 |====================================================================
b . 67.56 |====================================================================
Neural Magic DeepSparse 1.5
Model: ResNet-50, Sparse INT8 - Scenario: Asynchronous Multi-Stream
items/sec > Higher Is Better
a . 3227.10 |==================================================================
b . 3233.96 |==================================================================
Neural Magic DeepSparse 1.5
Model: ResNet-50, Sparse INT8 - Scenario: Asynchronous Multi-Stream
ms/batch < Lower Is Better
a . 4.9416 |===================================================================
b . 4.9312 |===================================================================
HeFFTe - Highly Efficient FFT for Exascale 2.3
Test: c2c - Backend: Stock - Precision: double - X Y Z: 512
GFLOP/s > Higher Is Better
a . 40.74 |====================================================================
b . 40.66 |====================================================================
Palabos 2.3
Grid Size: 500
Mega Site Updates Per Second > Higher Is Better
a . 300.28 |===================================================================
b . 300.86 |===================================================================
HeFFTe - Highly Efficient FFT for Exascale 2.3
Test: r2c - Backend: Stock - Precision: double - X Y Z: 256
GFLOP/s > Higher Is Better
a . 76.90 |====================================================================
b . 77.03 |====================================================================
HeFFTe - Highly Efficient FFT for Exascale 2.3
Test: c2c - Backend: FFTW - Precision: float - X Y Z: 512
GFLOP/s > Higher Is Better
a . 78.83 |====================================================================
b . 78.96 |====================================================================
OpenFOAM 10
Input: drivaerFastback, Medium Mesh Size - Mesh Time
Seconds < Lower Is Better
a . 144.70 |===================================================================
b . 144.94 |===================================================================
HeFFTe - Highly Efficient FFT for Exascale 2.3
Test: r2c - Backend: FFTW - Precision: float - X Y Z: 512
GFLOP/s > Higher Is Better
a . 141.41 |===================================================================
b . 141.19 |===================================================================
Neural Magic DeepSparse 1.5
Model: ResNet-50, Baseline - Scenario: Asynchronous Multi-Stream
items/sec > Higher Is Better
a . 479.79 |===================================================================
b . 480.52 |===================================================================
Laghos 3.1
Test: Sedov Blast Wave, ube_922_hex.mesh
Major Kernels Total Rate > Higher Is Better
a . 216.86 |===================================================================
b . 217.19 |===================================================================
Neural Magic DeepSparse 1.5
Model: ResNet-50, Baseline - Scenario: Asynchronous Multi-Stream
ms/batch < Lower Is Better
a . 33.33 |====================================================================
b . 33.28 |====================================================================
Blender 3.6
Blend File: BMW27 - Compute: CPU-Only
Seconds < Lower Is Better
a . 47.15 |====================================================================
b . 47.22 |====================================================================
HeFFTe - Highly Efficient FFT for Exascale 2.3
Test: r2c - Backend: Stock - Precision: float - X Y Z: 512
GFLOP/s > Higher Is Better
a . 137.54 |===================================================================
b . 137.74 |===================================================================
Stress-NG 0.15.10
Test: AVL Tree
Bogo Ops/s > Higher Is Better
a . 294.26 |===================================================================
b . 294.66 |===================================================================
Palabos 2.3
Grid Size: 100
Mega Site Updates Per Second > Higher Is Better
a . 235.19 |===================================================================
b . 234.87 |===================================================================
Stress-NG 0.15.10
Test: Floating Point
Bogo Ops/s > Higher Is Better
a . 10587.48 |=================================================================
b . 10601.10 |=================================================================
Liquid-DSP 1.6
Threads: 16 - Buffer Length: 256 - Filter Length: 32
samples/s > Higher Is Better
a . 557945000 |================================================================
b . 558655000 |================================================================
HeFFTe - Highly Efficient FFT for Exascale 2.3
Test: r2c - Backend: FFTW - Precision: double - X Y Z: 256
GFLOP/s > Higher Is Better
a . 72.29 |====================================================================
b . 72.20 |====================================================================
Stress-NG 0.15.10
Test: Malloc
Bogo Ops/s > Higher Is Better
a . 99373474.31 |==============================================================
b . 99251227.28 |==============================================================
Stress-NG 0.15.10
Test: Hash
Bogo Ops/s > Higher Is Better
a . 5577252.32 |===============================================================
b . 5583978.14 |===============================================================
Neural Magic DeepSparse 1.5
Model: CV Detection, YOLOv5s COCO - Scenario: Asynchronous Multi-Stream
ms/batch < Lower Is Better
a . 76.56 |====================================================================
b . 76.47 |====================================================================
High Performance Conjugate Gradient 3.1
X Y Z: 144 144 144 - RT: 60
GFLOP/s > Higher Is Better
a . 27.42 |====================================================================
b . 27.39 |====================================================================
Neural Magic DeepSparse 1.5
Model: NLP Text Classification, BERT base uncased SST2, Sparse INT8 - Scenario: Asynchronous Multi-Stream
items/sec > Higher Is Better
a . 1074.82 |==================================================================
b . 1075.96 |==================================================================
Neural Magic DeepSparse 1.5
Model: BERT-Large, NLP Question Answering, Sparse INT8 - Scenario: Asynchronous Multi-Stream
items/sec > Higher Is Better
a . 504.61 |===================================================================
b . 505.13 |===================================================================
Timed LLVM Compilation 16.0
Build System: Ninja
Seconds < Lower Is Better
a . 263.15 |===================================================================
b . 262.88 |===================================================================
Stress-NG 0.15.10
Test: Pthread
Bogo Ops/s > Higher Is Better
a . 136846.01 |================================================================
b . 136709.81 |================================================================
Blender 3.6
Blend File: Fishy Cat - Compute: CPU-Only
Seconds < Lower Is Better
a . 64.07 |====================================================================
b . 64.01 |====================================================================
HeFFTe - Highly Efficient FFT for Exascale 2.3
Test: c2c - Backend: FFTW - Precision: double - X Y Z: 512
GFLOP/s > Higher Is Better
a . 43.97 |====================================================================
b . 44.01 |====================================================================
OpenFOAM 10
Input: drivaerFastback, Medium Mesh Size - Execution Time
Seconds < Lower Is Better
a . 615.99 |===================================================================
b . 615.46 |===================================================================
Neural Magic DeepSparse 1.5
Model: NLP Text Classification, BERT base uncased SST2, Sparse INT8 - Scenario: Asynchronous Multi-Stream
ms/batch < Lower Is Better
a . 14.86 |====================================================================
b . 14.85 |====================================================================
Neural Magic DeepSparse 1.5
Model: BERT-Large, NLP Question Answering, Sparse INT8 - Scenario: Asynchronous Multi-Stream
ms/batch < Lower Is Better
a . 31.68 |====================================================================
b . 31.65 |====================================================================
Timed PHP Compilation 8.1.9
Time To Compile
Seconds < Lower Is Better
a . 42.35 |====================================================================
b . 42.38 |====================================================================
Stress-NG 0.15.10
Test: MEMFD
Bogo Ops/s > Higher Is Better
a . 549.94 |===================================================================
b . 549.55 |===================================================================
Liquid-DSP 1.6
Threads: 32 - Buffer Length: 256 - Filter Length: 32
samples/s > Higher Is Better
a . 847085000 |================================================================
b . 847675000 |================================================================
Stress-NG 0.15.10
Test: Context Switching
Bogo Ops/s > Higher Is Better
a . 2572801.75 |===============================================================
b . 2571092.69 |===============================================================
Neural Magic DeepSparse 1.5
Model: NLP Document Classification, oBERT base uncased on IMDB - Scenario: Asynchronous Multi-Stream
items/sec > Higher Is Better
a . 34.53 |====================================================================
b . 34.55 |====================================================================
Neural Magic DeepSparse 1.5
Model: CV Classification, ResNet-50 ImageNet - Scenario: Asynchronous Multi-Stream
items/sec > Higher Is Better
a . 478.91 |===================================================================
b . 479.22 |===================================================================
Neural Magic DeepSparse 1.5
Model: CV Classification, ResNet-50 ImageNet - Scenario: Asynchronous Multi-Stream
ms/batch < Lower Is Better
a . 33.39 |====================================================================
b . 33.37 |====================================================================
Stress-NG 0.15.10
Test: Poll
Bogo Ops/s > Higher Is Better
a . 3669281.69 |===============================================================
b . 3671617.97 |===============================================================
VVenC 1.9
Video Input: Bosphorus 1080p - Video Preset: Faster
Frames Per Second > Higher Is Better
a . 30.95 |====================================================================
b . 30.93 |====================================================================
Stress-NG 0.15.10
Test: Memory Copying
Bogo Ops/s > Higher Is Better
a . 7176.19 |==================================================================
b . 7180.43 |==================================================================
OpenFOAM 10
Input: drivaerFastback, Small Mesh Size - Mesh Time
Seconds < Lower Is Better
a . 27.97 |====================================================================
b . 27.95 |====================================================================
Stress-NG 0.15.10
Test: Matrix 3D Math
Bogo Ops/s > Higher Is Better
a . 9599.93 |==================================================================
b . 9605.30 |==================================================================
Stress-NG 0.15.10
Test: Forking
Bogo Ops/s > Higher Is Better
a . 89918.21 |=================================================================
b . 89966.29 |=================================================================
Neural Magic DeepSparse 1.5
Model: CV Detection, YOLOv5s COCO - Scenario: Asynchronous Multi-Stream
items/sec > Higher Is Better
a . 208.90 |===================================================================
b . 208.99 |===================================================================
Stress-NG 0.15.10
Test: Glibc Qsort Data Sorting
Bogo Ops/s > Higher Is Better
a . 696.65 |===================================================================
b . 696.92 |===================================================================
Stress-NG 0.15.10
Test: Zlib
Bogo Ops/s > Higher Is Better
a . 2647.81 |==================================================================
b . 2648.81 |==================================================================
Stress-NG 0.15.10
Test: System V Message Passing
Bogo Ops/s > Higher Is Better
a . 5852281.71 |===============================================================
b . 5854201.78 |===============================================================
Blender 3.6
Blend File: Barbershop - Compute: CPU-Only
Seconds < Lower Is Better
a . 493.45 |===================================================================
b . 493.61 |===================================================================
Timed Linux Kernel Compilation 6.1
Build: defconfig
Seconds < Lower Is Better
a . 40.44 |====================================================================
b . 40.45 |====================================================================
HeFFTe - Highly Efficient FFT for Exascale 2.3
Test: c2c - Backend: Stock - Precision: float - X Y Z: 512
GFLOP/s > Higher Is Better
a . 72.56 |====================================================================
b . 72.54 |====================================================================
Stress-NG 0.15.10
Test: Vector Math
Bogo Ops/s > Higher Is Better
a . 151386.31 |================================================================
b . 151431.15 |================================================================
Liquid-DSP 1.6
Threads: 64 - Buffer Length: 256 - Filter Length: 32
samples/s > Higher Is Better
a . 1577300000 |===============================================================
b . 1576850000 |===============================================================
Liquid-DSP 1.6
Threads: 64 - Buffer Length: 256 - Filter Length: 512
samples/s > Higher Is Better
a . 513135000 |================================================================
b . 513040000 |================================================================
Stress-NG 0.15.10
Test: Vector Floating Point
Bogo Ops/s > Higher Is Better
a . 58243.38 |=================================================================
b . 58232.70 |=================================================================
Blender 3.6
Blend File: Classroom - Compute: CPU-Only
Seconds < Lower Is Better
a . 127.78 |===================================================================
b . 127.76 |===================================================================
Stress-NG 0.15.10
Test: CPU Stress
Bogo Ops/s > Higher Is Better
a . 64111.11 |=================================================================
b . 64118.87 |=================================================================
HeFFTe - Highly Efficient FFT for Exascale 2.3
Test: r2c - Backend: Stock - Precision: double - X Y Z: 512
GFLOP/s > Higher Is Better
a . 76.61 |====================================================================
b . 76.60 |====================================================================
Stress-NG 0.15.10
Test: Crypto
Bogo Ops/s > Higher Is Better
a . 50240.09 |=================================================================
b . 50243.48 |=================================================================
Neural Magic DeepSparse 1.5
Model: NLP Document Classification, oBERT base uncased on IMDB - Scenario: Asynchronous Multi-Stream
ms/batch < Lower Is Better
a . 460.78 |===================================================================
b . 460.76 |===================================================================
Stress-NG 0.15.10
Test: x86_64 RdRand
Bogo Ops/s > Higher Is Better
a . 331416.52 |================================================================
b . 331423.04 |================================================================
Stress-NG 0.15.10
Test: Vector Shuffle
Bogo Ops/s > Higher Is Better
a . 167204.21 |================================================================
b . 167202.07 |================================================================
Timed Linux Kernel Compilation 6.1
Build: allmodconfig
Seconds < Lower Is Better
a . 445.39 |===================================================================
b . 445.38 |===================================================================
BRL-CAD 7.36
VGR Performance Metric
VGR Performance Metric > Higher Is Better
a . 466686 |===================================================================
Apache Cassandra 4.1.3
Test: Writes
Op/s > Higher Is Better
a . 155626 |===================================================================
Blender 3.6
Blend File: Pabellon Barcelona - Compute: CPU-Only
Seconds < Lower Is Better
a . 159.94 |===================================================================
Redis 7.0.12 + memtier_benchmark 2.0
Protocol: Redis - Clients: 500 - Set To Get Ratio: 1:10
Redis 7.0.12 + memtier_benchmark 2.0
Protocol: Redis - Clients: 500 - Set To Get Ratio: 1:5