GCC 11 vs. LLVM Clang 12 Benchmarks On Xeon Ice Lake Xeon Platinum 8380 compiler benchmarks by Michael Larabel looking at GCC 11 against LLVM Clang 12 for some initial holiday weekend tests... GCC 11.1: Processor: 2 x Intel Xeon Platinum 8380 @ 3.40GHz (80 Cores / 160 Threads), Motherboard: Intel M50CYP2SB2U (SE5C6200.86B.0022.D08.2103221623 BIOS), Chipset: Intel Device 0998, Memory: 16 x 32 GB DDR4-3200MT/s Hynix HMA84GR7CJR4N-XN, Disk: 800GB INTEL SSDPF21Q800GB, Graphics: ASPEED, Network: 2 x Intel X710 for 10GBASE-T + 2 x Intel E810-C for QSFP OS: Fedora 34, Kernel: 5.12.6-300.fc34.x86_64 (x86_64), Compiler: GCC 11.1.1 20210428, File-System: xfs, Screen Resolution: 1024x768 Clang 12.0: Processor: 2 x Intel Xeon Platinum 8380 @ 3.40GHz (80 Cores / 160 Threads), Motherboard: Intel M50CYP2SB2U (SE5C6200.86B.0022.D08.2103221623 BIOS), Chipset: Intel Device 0998, Memory: 16 x 32 GB DDR4-3200MT/s Hynix HMA84GR7CJR4N-XN, Disk: 800GB INTEL SSDPF21Q800GB, Graphics: ASPEED, Network: 2 x Intel X710 for 10GBASE-T + 2 x Intel E810-C for QSFP OS: Fedora 34, Kernel: 5.12.6-300.fc34.x86_64 (x86_64), Compiler: Clang 12.0.0, File-System: xfs, Screen Resolution: 1024x768 Crypto++ 8.2 Test: Unkeyed Algorithms MiB/second > Higher Is Better GCC 11.1 ... 359.90 |====================================================== Clang 12.0 . 388.57 |========================================================== Timed MrBayes Analysis 3.2.7 Primate Phylogeny Analysis Seconds < Lower Is Better GCC 11.1 ... 142.20 |========================================================== Clang 12.0 . 138.55 |========================================================= WebP Image Encode 1.1 Encode Settings: Default Encode Time - Seconds < Lower Is Better GCC 11.1 ... 1.638 |=========================================================== Clang 12.0 . 1.616 |========================================================== WebP Image Encode 1.1 Encode Settings: Quality 100 Encode Time - Seconds < Lower Is Better GCC 11.1 ... 2.645 |========================================================== Clang 12.0 . 2.692 |=========================================================== WebP Image Encode 1.1 Encode Settings: Quality 100, Lossless Encode Time - Seconds < Lower Is Better GCC 11.1 ... 19.47 |========================================================= Clang 12.0 . 20.27 |=========================================================== WebP Image Encode 1.1 Encode Settings: Quality 100, Highest Compression Encode Time - Seconds < Lower Is Better GCC 11.1 ... 8.026 |=========================================================== Clang 12.0 . 7.422 |======================================================= WebP Image Encode 1.1 Encode Settings: Quality 100, Lossless, Highest Compression Encode Time - Seconds < Lower Is Better GCC 11.1 ... 40.91 |========================================================= Clang 12.0 . 42.18 |=========================================================== GNU GMP GMPbench 6.2.1 Total Time GMPbench Score > Higher Is Better GCC 11.1 . 3871.6 |============================================================ Zstd Compression 1.5.0 Compression Level: 8 - Compression Speed MB/s > Higher Is Better GCC 11.1 ... 2611.0 |======================================================= Clang 12.0 . 2748.1 |========================================================== Zstd Compression 1.5.0 Compression Level: 8 - Decompression Speed MB/s > Higher Is Better GCC 11.1 ... 2959.3 |========================================================= Clang 12.0 . 2996.7 |========================================================== Zstd Compression 1.5.0 Compression Level: 19 - Compression Speed MB/s > Higher Is Better GCC 11.1 ... 83.8 |============================================================ Clang 12.0 . 81.5 |========================================================== Zstd Compression 1.5.0 Compression Level: 19 - Decompression Speed MB/s > Higher Is Better GCC 11.1 ... 2537.3 |========================================================== Clang 12.0 . 2495.2 |========================================================= Zstd Compression 1.5.0 Compression Level: 8, Long Mode - Compression Speed MB/s > Higher Is Better GCC 11.1 ... 1040.6 |========================================================== Clang 12.0 . 830.0 |============================================== Zstd Compression 1.5.0 Compression Level: 8, Long Mode - Decompression Speed MB/s > Higher Is Better GCC 11.1 ... 3168.4 |========================================================= Clang 12.0 . 3204.6 |========================================================== Zstd Compression 1.5.0 Compression Level: 19, Long Mode - Compression Speed MB/s > Higher Is Better GCC 11.1 ... 47.9 |============================================================ Clang 12.0 . 46.2 |========================================================== Zstd Compression 1.5.0 Compression Level: 19, Long Mode - Decompression Speed MB/s > Higher Is Better GCC 11.1 ... 2670.8 |========================================================== Clang 12.0 . 2632.3 |========================================================= GraphicsMagick 1.3.33 Operation: Rotate Iterations Per Minute > Higher Is Better GCC 11.1 ... 745 |============================================================= Clang 12.0 . 745 |============================================================= GraphicsMagick 1.3.33 Operation: Sharpen Iterations Per Minute > Higher Is Better GCC 11.1 ... 898 |============================================================= Clang 12.0 . 769 |==================================================== GraphicsMagick 1.3.33 Operation: Enhanced Iterations Per Minute > Higher Is Better GCC 11.1 ... 1315 |============================================================ Clang 12.0 . 1141 |==================================================== GraphicsMagick 1.3.33 Operation: Resizing Iterations Per Minute > Higher Is Better GCC 11.1 ... 380 |====================================== Clang 12.0 . 614 |============================================================= Kvazaar 2.0 Video Input: Bosphorus 4K - Video Preset: Very Fast Frames Per Second > Higher Is Better GCC 11.1 ... 42.31 |======================================================== Clang 12.0 . 44.72 |=========================================================== Kvazaar 2.0 Video Input: Bosphorus 4K - Video Preset: Ultra Fast Frames Per Second > Higher Is Better GCC 11.1 ... 46.58 |========================================================= Clang 12.0 . 47.96 |=========================================================== Kvazaar 2.0 Video Input: Bosphorus 1080p - Video Preset: Very Fast Frames Per Second > Higher Is Better GCC 11.1 ... 159.71 |======================================================== Clang 12.0 . 166.00 |========================================================== Kvazaar 2.0 Video Input: Bosphorus 1080p - Video Preset: Ultra Fast Frames Per Second > Higher Is Better GCC 11.1 ... 176.66 |======================================================== Clang 12.0 . 183.97 |========================================================== SVT-AV1 0.8.7 Encoder Mode: Preset 4 - Input: Bosphorus 4K Frames Per Second > Higher Is Better GCC 11.1 ... 4.213 |========================================================= Clang 12.0 . 4.333 |=========================================================== SVT-AV1 0.8.7 Encoder Mode: Preset 8 - Input: Bosphorus 4K Frames Per Second > Higher Is Better GCC 11.1 ... 55.19 |========================================================== Clang 12.0 . 56.27 |=========================================================== SVT-AV1 0.8.7 Encoder Mode: Preset 4 - Input: Bosphorus 1080p Frames Per Second > Higher Is Better GCC 11.1 ... 8.996 |======================================================== Clang 12.0 . 9.418 |=========================================================== SVT-AV1 0.8.7 Encoder Mode: Preset 8 - Input: Bosphorus 1080p Frames Per Second > Higher Is Better GCC 11.1 ... 167.39 |========================================================= Clang 12.0 . 169.61 |========================================================== SVT-HEVC 1.5.0 Tuning: 1 - Input: Bosphorus 1080p Frames Per Second > Higher Is Better GCC 11.1 ... 39.59 |======================================================== Clang 12.0 . 41.96 |=========================================================== SVT-HEVC 1.5.0 Tuning: 7 - Input: Bosphorus 1080p Frames Per Second > Higher Is Better GCC 11.1 ... 336.37 |======================================================= Clang 12.0 . 355.18 |========================================================== SVT-HEVC 1.5.0 Tuning: 10 - Input: Bosphorus 1080p Frames Per Second > Higher Is Better GCC 11.1 ... 609.56 |========================================================== Clang 12.0 . 608.95 |========================================================== SVT-VP9 0.3 Tuning: VMAF Optimized - Input: Bosphorus 1080p Frames Per Second > Higher Is Better GCC 11.1 ... 476.85 |========================================================== Clang 12.0 . 467.57 |========================================================= SVT-VP9 0.3 Tuning: PSNR/SSIM Optimized - Input: Bosphorus 1080p Frames Per Second > Higher Is Better GCC 11.1 ... 477.90 |========================================================== Clang 12.0 . 466.06 |========================================================= SVT-VP9 0.3 Tuning: Visual Quality Optimized - Input: Bosphorus 1080p Frames Per Second > Higher Is Better GCC 11.1 ... 393.17 |========================================================== Clang 12.0 . 379.93 |======================================================== x265 3.4 Video Input: Bosphorus 4K Frames Per Second > Higher Is Better GCC 11.1 ... 26.92 |======================================================= Clang 12.0 . 29.13 |=========================================================== x265 3.4 Video Input: Bosphorus 1080p Frames Per Second > Higher Is Better GCC 11.1 ... 76.88 |=========================================================== Clang 12.0 . 77.24 |=========================================================== Coremark 1.0 CoreMark Size 666 - Iterations Per Second Iterations/Sec > Higher Is Better GCC 11.1 ... 2522898.57 |====================================================== Clang 12.0 . 2130829.93 |============================================== Himeno Benchmark 3.0 Poisson Pressure Solver MFLOPS > Higher Is Better GCC 11.1 ... 4651.53 |========================================================= Clang 12.0 . 4161.65 |=================================================== C-Ray 1.1 Total Time - 4K, 16 Rays Per Pixel Seconds < Lower Is Better GCC 11.1 ... 7.794 |============================= Clang 12.0 . 15.352 |========================================================== Primesieve 7.4 1e12 Prime Number Generation Seconds < Lower Is Better GCC 11.1 ... 3.780 |========================================================== Clang 12.0 . 3.830 |=========================================================== oneDNN 2.1.2 Harness: IP Shapes 1D - Data Type: f32 - Engine: CPU ms < Lower Is Better GCC 11.1 ... 0.923876 |======================================================== Clang 12.0 . 0.665355 |======================================== oneDNN 2.1.2 Harness: IP Shapes 3D - Data Type: f32 - Engine: CPU ms < Lower Is Better GCC 11.1 ... 1.40268 |========================================================= Clang 12.0 . 1.21726 |================================================= oneDNN 2.1.2 Harness: IP Shapes 1D - Data Type: u8s8f32 - Engine: CPU ms < Lower Is Better GCC 11.1 ... 1.23707 |========================================================= Clang 12.0 . 1.00883 |============================================== oneDNN 2.1.2 Harness: IP Shapes 3D - Data Type: u8s8f32 - Engine: CPU ms < Lower Is Better GCC 11.1 ... 0.439708 |======================================================== Clang 12.0 . 0.325027 |========================================= oneDNN 2.1.2 Harness: IP Shapes 1D - Data Type: bf16bf16bf16 - Engine: CPU ms < Lower Is Better GCC 11.1 ... 3.00252 |========================================================= Clang 12.0 . 2.76681 |===================================================== oneDNN 2.1.2 Harness: IP Shapes 3D - Data Type: bf16bf16bf16 - Engine: CPU ms < Lower Is Better GCC 11.1 ... 1.81351 |========================================================= Clang 12.0 . 1.70567 |====================================================== oneDNN 2.1.2 Harness: Convolution Batch Shapes Auto - Data Type: f32 - Engine: CPU ms < Lower Is Better GCC 11.1 ... 1.45930 |========================================================= Clang 12.0 . 1.41167 |======================================================= oneDNN 2.1.2 Harness: Deconvolution Batch shapes_3d - Data Type: f32 - Engine: CPU ms < Lower Is Better GCC 11.1 ... 0.840383 |======================================================== Clang 12.0 . 0.842984 |======================================================== oneDNN 2.1.2 Harness: Convolution Batch Shapes Auto - Data Type: u8s8f32 - Engine: CPU ms < Lower Is Better GCC 11.1 ... 0.969972 |======================================================== Clang 12.0 . 0.902213 |==================================================== oneDNN 2.1.2 Harness: Deconvolution Batch shapes_1d - Data Type: u8s8f32 - Engine: CPU ms < Lower Is Better GCC 11.1 ... 0.361691 |======================================================== Clang 12.0 . 0.200917 |=============================== oneDNN 2.1.2 Harness: Deconvolution Batch shapes_3d - Data Type: u8s8f32 - Engine: CPU ms < Lower Is Better GCC 11.1 ... 0.194587 |======================================================== Clang 12.0 . 0.175060 |================================================== oneDNN 2.1.2 Harness: Recurrent Neural Network Training - Data Type: f32 - Engine: CPU ms < Lower Is Better GCC 11.1 ... 686.19 |========================================================== Clang 12.0 . 589.28 |================================================== oneDNN 2.1.2 Harness: Recurrent Neural Network Inference - Data Type: f32 - Engine: CPU ms < Lower Is Better GCC 11.1 ... 448.25 |========================================================== Clang 12.0 . 352.47 |============================================== oneDNN 2.1.2 Harness: Recurrent Neural Network Training - Data Type: u8s8f32 - Engine: CPU ms < Lower Is Better GCC 11.1 ... 686.05 |========================================================== Clang 12.0 . 594.11 |================================================== oneDNN 2.1.2 Harness: Convolution Batch Shapes Auto - Data Type: bf16bf16bf16 - Engine: CPU ms < Lower Is Better GCC 11.1 ... 2.09246 |========================================================= Clang 12.0 . 2.06629 |======================================================== oneDNN 2.1.2 Harness: Deconvolution Batch shapes_1d - Data Type: bf16bf16bf16 - Engine: CPU ms < Lower Is Better GCC 11.1 ... 3.27829 |========================================================= Clang 12.0 . 2.86766 |================================================== oneDNN 2.1.2 Harness: Deconvolution Batch shapes_3d - Data Type: bf16bf16bf16 - Engine: CPU ms < Lower Is Better GCC 11.1 ... 3.57127 |========================================================= Clang 12.0 . 3.57288 |========================================================= oneDNN 2.1.2 Harness: Recurrent Neural Network Inference - Data Type: u8s8f32 - Engine: CPU ms < Lower Is Better GCC 11.1 ... 445.26 |========================================================== Clang 12.0 . 352.59 |============================================== oneDNN 2.1.2 Harness: Matrix Multiply Batch Shapes Transformer - Data Type: f32 - Engine: CPU ms < Lower Is Better GCC 11.1 ... 0.247554 |======================================================== Clang 12.0 . 0.160017 |==================================== oneDNN 2.1.2 Harness: Recurrent Neural Network Training - Data Type: bf16bf16bf16 - Engine: CPU ms < Lower Is Better GCC 11.1 ... 677.71 |========================================================== Clang 12.0 . 584.54 |================================================== oneDNN 2.1.2 Harness: Recurrent Neural Network Inference - Data Type: bf16bf16bf16 - Engine: CPU ms < Lower Is Better GCC 11.1 ... 444.23 |========================================================== Clang 12.0 . 353.58 |============================================== oneDNN 2.1.2 Harness: Matrix Multiply Batch Shapes Transformer - Data Type: u8s8f32 - Engine: CPU ms < Lower Is Better GCC 11.1 ... 0.219899 |======================================================== Clang 12.0 . 0.120406 |=============================== oneDNN 2.1.2 Harness: Matrix Multiply Batch Shapes Transformer - Data Type: bf16bf16bf16 - Engine: CPU ms < Lower Is Better GCC 11.1 ... 0.601530 |======================================================== Clang 12.0 . 0.515284 |================================================ AOBench Size: 2048 x 2048 - Total Time Seconds < Lower Is Better GCC 11.1 ... 33.88 |======================================================= Clang 12.0 . 36.08 |=========================================================== Bullet Physics Engine 2.81 Test: 3000 Fall Seconds < Lower Is Better GCC 11.1 ... 3.873060 |================================================== Clang 12.0 . 4.301823 |======================================================== Bullet Physics Engine 2.81 Test: 1000 Stack Seconds < Lower Is Better GCC 11.1 ... 4.451765 |============================================== Clang 12.0 . 5.410511 |======================================================== Bullet Physics Engine 2.81 Test: 1000 Convex Seconds < Lower Is Better GCC 11.1 ... 4.340600 |=================================================== Clang 12.0 . 4.741088 |======================================================== Bullet Physics Engine 2.81 Test: 136 Ragdolls Seconds < Lower Is Better GCC 11.1 ... 2.553303 |=================================================== Clang 12.0 . 2.798322 |======================================================== Bullet Physics Engine 2.81 Test: Prim Trimesh Seconds < Lower Is Better GCC 11.1 ... 0.862000 |=================================================== Clang 12.0 . 0.940903 |======================================================== Bullet Physics Engine 2.81 Test: Convex Trimesh Seconds < Lower Is Better GCC 11.1 ... 1.054273 |=================================================== Clang 12.0 . 1.148367 |======================================================== FLAC Audio Encoding 1.3.2 WAV To FLAC Seconds < Lower Is Better GCC 11.1 ... 9.382 |========================================================== Clang 12.0 . 9.474 |=========================================================== LAME MP3 Encoding 3.100 WAV To MP3 Seconds < Lower Is Better GCC 11.1 ... 8.619 |===================================================== Clang 12.0 . 9.591 |=========================================================== Opus Codec Encoding 1.3.1 WAV To Opus Encode Seconds < Lower Is Better GCC 11.1 ... 8.768 |===================================================== Clang 12.0 . 9.788 |=========================================================== eSpeak-NG Speech Engine 20200907 Text-To-Speech Synthesis Seconds < Lower Is Better GCC 11.1 ... 30.51 |=========================================================== Clang 12.0 . 29.81 |========================================================== Gcrypt Library 1.9 Seconds < Lower Is Better GCC 11.1 ... 265.17 |========================================================== Clang 12.0 . 253.86 |======================================================== WebP2 Image Encode 20210126 Encode Settings: Default Seconds < Lower Is Better GCC 11.1 ... 2.644 |=========================================================== Clang 12.0 . 2.492 |======================================================== WebP2 Image Encode 20210126 Encode Settings: Quality 75, Compression Effort 7 Seconds < Lower Is Better GCC 11.1 ... 106.66 |========================================================== Clang 12.0 . 99.78 |====================================================== WebP2 Image Encode 20210126 Encode Settings: Quality 95, Compression Effort 7 Seconds < Lower Is Better GCC 11.1 ... 196.49 |========================================================== Clang 12.0 . 182.29 |====================================================== WebP2 Image Encode 20210126 Encode Settings: Quality 100, Compression Effort 5 Seconds < Lower Is Better GCC 11.1 ... 5.765 |==================================================== Clang 12.0 . 6.495 |=========================================================== WebP2 Image Encode 20210126 Encode Settings: Quality 100, Lossless Compression Seconds < Lower Is Better GCC 11.1 ... 389.07 |======================================================== Clang 12.0 . 400.62 |========================================================== OpenSSL 1.1.1 RSA 4096-bit Performance Signs Per Second > Higher Is Better GCC 11.1 ... 17804.2 |========================================================= Clang 12.0 . 11555.8 |===================================== Liquid-DSP 2021.01.31 Threads: 1 - Buffer Length: 256 - Filter Length: 57 samples/s > Higher Is Better GCC 11.1 ... 60985333 |======================================================= Clang 12.0 . 61840333 |======================================================== Liquid-DSP 2021.01.31 Threads: 160 - Buffer Length: 256 - Filter Length: 57 samples/s > Higher Is Better GCC 11.1 ... 3182866667 |=============================================== Clang 12.0 . 3686466667 |====================================================== libjpeg-turbo tjbench 2.1.0 Test: Decompression Throughput Megapixels/sec > Higher Is Better GCC 11.1 ... 174.23 |======================================================== Clang 12.0 . 180.60 |========================================================== Darmstadt Automotive Parallel Heterogeneous Suite Backend: OpenMP - Kernel: NDT Mapping Test Cases Per Minute > Higher Is Better GCC 11.1 . 1046.60 |=========================================================== Darmstadt Automotive Parallel Heterogeneous Suite Backend: OpenMP - Kernel: Points2Image Test Cases Per Minute > Higher Is Better GCC 11.1 . 14507.81 |========================================================== Darmstadt Automotive Parallel Heterogeneous Suite Backend: OpenMP - Kernel: Euclidean Cluster Test Cases Per Minute > Higher Is Better GCC 11.1 . 1013.99 |=========================================================== PostgreSQL pgbench 13.0 Scaling Factor: 100 - Clients: 250 - Mode: Read Only TPS > Higher Is Better GCC 11.1 ... 907401 |======================================================== Clang 12.0 . 943043 |========================================================== PostgreSQL pgbench 13.0 Scaling Factor: 100 - Clients: 250 - Mode: Read Only - Average Latency ms < Lower Is Better GCC 11.1 ... 0.277 |=========================================================== Clang 12.0 . 0.265 |======================================================== PostgreSQL pgbench 13.0 Scaling Factor: 100 - Clients: 250 - Mode: Read Write TPS > Higher Is Better GCC 11.1 ... 89425 |========================================================= Clang 12.0 . 92576 |=========================================================== PostgreSQL pgbench 13.0 Scaling Factor: 100 - Clients: 250 - Mode: Read Write - Average Latency ms < Lower Is Better GCC 11.1 ... 2.797 |=========================================================== Clang 12.0 . 2.702 |========================================================= ASTC Encoder 2.4 Preset: Medium Seconds < Lower Is Better GCC 11.1 ... 6.4270 |========================================================== Clang 12.0 . 4.5120 |========================================= ASTC Encoder 2.4 Preset: Thorough Seconds < Lower Is Better GCC 11.1 ... 9.4219 |========================================================== Clang 12.0 . 6.9380 |=========================================== ASTC Encoder 2.4 Preset: Exhaustive Seconds < Lower Is Better GCC 11.1 ... 16.39 |=========================================================== Clang 12.0 . 14.01 |================================================== Caffe 2020-02-13 Model: AlexNet - Acceleration: CPU - Iterations: 200 Milli-Seconds < Lower Is Better GCC 11.1 ... 298291 |========================================================== Clang 12.0 . 297554 |========================================================== Caffe 2020-02-13 Model: GoogleNet - Acceleration: CPU - Iterations: 200 Milli-Seconds < Lower Is Better GCC 11.1 ... 662408 |========================================================== Clang 12.0 . 663282 |========================================================== NCNN 20201218 Target: CPU - Model: mobilenet ms < Lower Is Better GCC 11.1 ... 19.40 |=========================================================== Clang 12.0 . 14.21 |=========================================== NCNN 20201218 Target: CPU-v2-v2 - Model: mobilenet-v2 ms < Lower Is Better GCC 11.1 ... 9.80 |============================================================ Clang 12.0 . 4.95 |============================== NCNN 20201218 Target: CPU-v3-v3 - Model: mobilenet-v3 ms < Lower Is Better GCC 11.1 ... 9.56 |============================================================ Clang 12.0 . 4.22 |========================== NCNN 20201218 Target: CPU - Model: shufflenet-v2 ms < Lower Is Better GCC 11.1 ... 10.55 |=========================================================== Clang 12.0 . 5.90 |================================= NCNN 20201218 Target: CPU - Model: mnasnet ms < Lower Is Better GCC 11.1 ... 9.43 |============================================================ Clang 12.0 . 5.26 |================================= NCNN 20201218 Target: CPU - Model: efficientnet-b0 ms < Lower Is Better GCC 11.1 ... 12.48 |=========================================================== Clang 12.0 . 7.72 |==================================== NCNN 20201218 Target: CPU - Model: blazeface ms < Lower Is Better GCC 11.1 ... 6.15 |============================================================ Clang 12.0 . 2.38 |======================= NCNN 20201218 Target: CPU - Model: googlenet ms < Lower Is Better GCC 11.1 ... 19.46 |=========================================================== Clang 12.0 . 15.34 |=============================================== NCNN 20201218 Target: CPU - Model: vgg16 ms < Lower Is Better GCC 11.1 ... 25.34 |======================================================= Clang 12.0 . 27.34 |=========================================================== NCNN 20201218 Target: CPU - Model: resnet18 ms < Lower Is Better GCC 11.1 ... 11.10 |=========================================================== Clang 12.0 . 10.83 |========================================================== NCNN 20201218 Target: CPU - Model: yolov4-tiny ms < Lower Is Better GCC 11.1 ... 23.12 |====================================================== Clang 12.0 . 25.28 |=========================================================== NCNN 20201218 Target: CPU - Model: squeezenet_ssd ms < Lower Is Better GCC 11.1 ... 21.18 |=========================================================== Clang 12.0 . 17.96 |================================================== NCNN 20201218 Target: CPU - Model: regnety_400m ms < Lower Is Better GCC 11.1 ... 94.38 |=========================================================== Clang 12.0 . 26.47 |================= TNN 0.2.3 Target: CPU - Model: MobileNet v2 ms < Lower Is Better GCC 11.1 ... 376.74 |======================================== Clang 12.0 . 552.03 |========================================================== TNN 0.2.3 Target: CPU - Model: SqueezeNet v1.1 ms < Lower Is Better GCC 11.1 ... 377.43 |====================================================== Clang 12.0 . 402.91 |========================================================== WavPack Audio Encoding 5.3 WAV To WavPack Seconds < Lower Is Better GCC 11.1 ... 17.36 |=========================================================== Clang 12.0 . 17.34 |=========================================================== Kripke 1.2.4 Throughput FoM > Higher Is Better GCC 11.1 ... 177613600 |======================================================= Clang 12.0 . 160155675 |================================================== Geometric Mean Of All Test Results Result Composite - GCC 11 vs. LLVM Clang 12 Benchmarks On Xeon Ice Lake Geometric Mean > Higher Is Better GCC 11.1 ... 50.57 |====================================================== Clang 12.0 . 55.01 |===========================================================