new tests may aarch64
Ampere ARMv8 Neoverse-N1 testing with a WIWYNN Mt.Jade (2.03.20210719 SCP: BIOS) and ASPEED on Ubuntu 21.10 via the Phoronix Test Suite.
HTML result view exported from: https://openbenchmarking.org/result/2205067-NE-NEWTESTSM45&grs&sor.
oneDNN
Harness: Convolution Batch Shapes Auto - Data Type: f32 - Engine: CPU
oneDNN
Harness: Matrix Multiply Batch Shapes Transformer - Data Type: u8s8f32 - Engine: CPU
oneDNN
Harness: Recurrent Neural Network Inference - Data Type: u8s8f32 - Engine: CPU
oneDNN
Harness: Recurrent Neural Network Training - Data Type: u8s8f32 - Engine: CPU
ONNX Runtime
Model: GPT-2 - Device: CPU - Executor: Standard
oneDNN
Harness: Convolution Batch Shapes Auto - Data Type: u8s8f32 - Engine: CPU
oneDNN
Harness: Recurrent Neural Network Training - Data Type: bf16bf16bf16 - Engine: CPU
Stress-NG
Test: SENDFILE
Apache HTTP Server
Concurrent Requests: 1
ONNX Runtime
Model: ArcFace ResNet-100 - Device: CPU - Executor: Standard
oneDNN
Harness: Recurrent Neural Network Inference - Data Type: bf16bf16bf16 - Engine: CPU
Apache HTTP Server
Concurrent Requests: 200
oneDNN
Harness: Deconvolution Batch shapes_1d - Data Type: f32 - Engine: CPU
oneDNN
Harness: Matrix Multiply Batch Shapes Transformer - Data Type: f32 - Engine: CPU
Apache HTTP Server
Concurrent Requests: 1000
oneDNN
Harness: Recurrent Neural Network Inference - Data Type: f32 - Engine: CPU
Apache HTTP Server
Concurrent Requests: 20
perf-bench
Benchmark: Memcpy 1MB
Stress-NG
Test: CPU Cache
Stress-NG
Test: Futex
perf-bench
Benchmark: Sched Pipe
perf-bench
Benchmark: Epoll Wait
oneDNN
Harness: Deconvolution Batch shapes_1d - Data Type: u8s8f32 - Engine: CPU
oneDNN
Harness: IP Shapes 3D - Data Type: u8s8f32 - Engine: CPU
ONNX Runtime
Model: yolov4 - Device: CPU - Executor: Standard
Stress-NG
Test: MMAP
Stress-NG
Test: Forking
Stress-NG
Test: Malloc
Apache HTTP Server
Concurrent Requests: 100
libavif avifenc
Encoder Speed: 10, Lossless
Java JMH
Throughput
ONNX Runtime
Model: fcn-resnet101-11 - Device: CPU - Executor: Standard
oneDNN
Harness: Recurrent Neural Network Training - Data Type: f32 - Engine: CPU
oneDNN
Harness: Deconvolution Batch shapes_3d - Data Type: f32 - Engine: CPU
perf-bench
Benchmark: Futex Lock-Pi
ONNX Runtime
Model: bertsquad-12 - Device: CPU - Executor: Standard
Apache HTTP Server
Concurrent Requests: 500
ONNX Runtime
Model: GPT-2 - Device: CPU - Executor: Parallel
Stress-NG
Test: Memory Copying
oneDNN
Harness: IP Shapes 1D - Data Type: f32 - Engine: CPU
nginx
Concurrent Requests: 1
Stress-NG
Test: IO_uring
ONNX Runtime
Model: bertsquad-12 - Device: CPU - Executor: Parallel
oneDNN
Harness: Deconvolution Batch shapes_3d - Data Type: u8s8f32 - Engine: CPU
ONNX Runtime
Model: yolov4 - Device: CPU - Executor: Parallel
nginx
Concurrent Requests: 500
Stress-NG
Test: Context Switching
ONNX Runtime
Model: ArcFace ResNet-100 - Device: CPU - Executor: Parallel
ONNX Runtime
Model: super-resolution-10 - Device: CPU - Executor: Standard
libavif avifenc
Encoder Speed: 2
Stress-NG
Test: System V Message Passing
oneDNN
Harness: IP Shapes 1D - Data Type: u8s8f32 - Engine: CPU
Stress-NG
Test: Crypto
libavif avifenc
Encoder Speed: 6
WebP2 Image Encode
Encode Settings: Default
nginx
Concurrent Requests: 200
nginx
Concurrent Requests: 1000
oneDNN
Harness: IP Shapes 3D - Data Type: f32 - Engine: CPU
libavif avifenc
Encoder Speed: 6, Lossless
nginx
Concurrent Requests: 100
Stress-NG
Test: MEMFD
Stress-NG
Test: NUMA
Stress-NG
Test: Socket Activity
Stress-NG
Test: Glibc C String Functions
perf-bench
Benchmark: Syscall Basic
nginx
Concurrent Requests: 20
WebP2 Image Encode
Encode Settings: Quality 95, Compression Effort 7
libavif avifenc
Encoder Speed: 0
Stress-NG
Test: Glibc Qsort Data Sorting
WebP2 Image Encode
Encode Settings: Quality 75, Compression Effort 7
WebP2 Image Encode
Encode Settings: Quality 100, Lossless Compression
Stress-NG
Test: Semaphores
ONNX Runtime
Model: super-resolution-10 - Device: CPU - Executor: Parallel
WebP2 Image Encode
Encode Settings: Quality 100, Compression Effort 5
Stress-NG
Test: Matrix Math
Stress-NG
Test: Vector Math
Stress-NG
Test: CPU Stress
perf-bench
Benchmark: Futex Hash
perf-bench
Benchmark: Memset 1MB
ONNX Runtime
Model: fcn-resnet101-11 - Device: CPU - Executor: Parallel
Stress-NG
Test: Atomic
Phoronix Test Suite v10.8.4