Microsoft Azure HBv4 HPC Performance Benchmarks
Benchmarks for a future article on Phoronix looking at HBv4 Genoa-X Linux performance..


HC: 

	Processor: 2 x Intel Xeon Platinum 8168 (44 Cores), Motherboard: Microsoft Virtual Machine (Hyper-V UEFI v4.1 BIOS), Memory: 1 GB + 60928 MB + 118272 MB + 176 GB, Disk: 32GB Virtual Disk + 752GB Virtual Disk, Graphics: hyperv_fb

	OS: AlmaLinux 8.8, Kernel: 4.18.0-425.3.1.el8.x86_64 (x86_64), Compiler: GCC 13.1.0 + CUDA 12.1, File-System: nfs, Screen Resolution: 1024x768, System Layer: microsoft

HBv2: 

	Processor: 2 x AMD EPYC 7V12 64-Core (120 Cores), Motherboard: Microsoft Virtual Machine (Hyper-V UEFI v4.1 BIOS), Memory: 1 GB + 59 GB + 54 GB + 114 GB + 114 GB + 114 GB, Disk: 960GB Microsoft NVMe Direct Disk + 32GB Virtual Disk + 515GB Virtual Disk, Graphics: hyperv_fb

	OS: AlmaLinux 8.8, Kernel: 4.18.0-425.3.1.el8.x86_64 (x86_64), Compiler: GCC 13.1.0 + CUDA 12.1, File-System: nfs, Screen Resolution: 1024x768, System Layer: microsoft

HBv3: 

	Processor: 2 x AMD EPYC 7V73X 64-Core (120 Cores), Motherboard: Microsoft Virtual Machine (Hyper-V UEFI v4.1 BIOS), Memory: 1 GB + 59 GB + 54 GB + 114 GB + 114 GB + 114 GB, Disk: 2 x 960GB Microsoft NVMe Direct Disk + 32GB Virtual Disk + 515GB Virtual Disk, Graphics: hyperv_fb

	OS: AlmaLinux 8.8, Kernel: 4.18.0-425.3.1.el8.x86_64 (x86_64), Compiler: GCC 13.1.0 + CUDA 12.1, File-System: nfs, Screen Resolution: 1024x768, System Layer: microsoft

HBv4: 

	Processor: 2 x AMD EPYC 9V33X 96-Core (176 Cores), Motherboard: Microsoft Virtual Machine (Hyper-V UEFI v4.1 BIOS), Memory: 1 GB + 59 GB + 116 GB + 176 GB + 176 GB + 176 GB, Disk: 2 x 1920GB Microsoft NVMe Direct Disk + 32GB Virtual Disk + 515GB Virtual Disk, Graphics: hyperv_fb

	OS: AlmaLinux 8.8, Kernel: 4.18.0-425.3.1.el8.x86_64 (x86_64), Compiler: GCC 13.1.0 + CUDA 12.1, File-System: nfs, Screen Resolution: 1024x768, System Layer: microsoft


7-Zip Compression 22.01
Test: Compression Rating
MIPS > Higher Is Better
HC ... 216451  |=============
HBv2 . 501534  |=============================
HBv3 . 566595  |=================================
HBv4 . 1083523 |===============================================================


7-Zip Compression 22.01
Test: Decompression Rating
MIPS > Higher Is Better
HC ... 150841 |=============
HBv2 . 388577 |=================================
HBv3 . 406516 |===================================
HBv4 . 742859 |================================================================


ACES DGEMM 1.0
Sustained Floating-Point Rate
GFLOP/s > Higher Is Better
HC ... 14.072027 |================
HBv2 . 6.395415  |=======
HBv3 . 25.048352 |=============================
HBv4 . 52.802440 |=============================================================


Blender 3.6
Blend File: BMW27 - Compute: CPU-Only
Seconds < Lower Is Better
HC ... 49.95 |=================================================================
HBv2 . 19.58 |=========================
HBv3 . 19.43 |=========================
HBv4 . 10.11 |=============


Blender 3.6
Blend File: Classroom - Compute: CPU-Only
Seconds < Lower Is Better
HC ... 138.51 |================================================================
HBv2 . 50.95  |========================
HBv3 . 50.71  |=======================
HBv4 . 25.61  |============


Blender 3.6
Blend File: Fishy Cat - Compute: CPU-Only
Seconds < Lower Is Better
HC ... 71.76 |=================================================================
HBv2 . 26.43 |========================
HBv3 . 25.59 |=======================
HBv4 . 13.74 |============


Blender 3.6
Blend File: Barbershop - Compute: CPU-Only
Seconds < Lower Is Better
HC ... 526.93 |================================================================
HBv2 . 211.46 |==========================
HBv3 . 188.96 |=======================
HBv4 . 97.52  |============


Blender 3.6
Blend File: Pabellon Barcelona - Compute: CPU-Only
Seconds < Lower Is Better
HC ... 175.07 |================================================================
HBv2 . 64.84  |========================
HBv3 . 62.90  |=======================
HBv4 . 33.01  |============


HeFFTe - Highly Efficient FFT for Exascale 2.3
Test: c2c - Backend: FFTW - Precision: float - X Y Z: 256
GFLOP/s > Higher Is Better
HC ... 58.36  |===============
HBv2 . 91.54  |=======================
HBv3 . 103.51 |==========================
HBv4 . 256.35 |================================================================


HeFFTe - Highly Efficient FFT for Exascale 2.3
Test: c2c - Backend: FFTW - Precision: float - X Y Z: 512
GFLOP/s > Higher Is Better
HC ... 62.98  |===========
HBv2 . 95.88  |=================
HBv3 . 135.69 |========================
HBv4 . 355.86 |================================================================


HeFFTe - Highly Efficient FFT for Exascale 2.3
Test: r2c - Backend: FFTW - Precision: float - X Y Z: 512
GFLOP/s > Higher Is Better
HC ... 114.03 |============
HBv2 . 191.78 |====================
HBv3 . 254.25 |==========================
HBv4 . 622.58 |================================================================


HeFFTe - Highly Efficient FFT for Exascale 2.3
Test: c2c - Backend: FFTW - Precision: double - X Y Z: 512
GFLOP/s > Higher Is Better
HC ... 33.52  |=============
HBv2 . 47.61  |===================
HBv3 . 57.33  |=======================
HBv4 . 159.18 |================================================================


HeFFTe - Highly Efficient FFT for Exascale 2.3
Test: c2c - Backend: Stock - Precision: float - X Y Z: 256
GFLOP/s > Higher Is Better
HC ... 59.73  |================
HBv2 . 91.26  |========================
HBv3 . 103.41 |===========================
HBv4 . 244.34 |================================================================


HeFFTe - Highly Efficient FFT for Exascale 2.3
Test: c2c - Backend: Stock - Precision: float - X Y Z: 512
GFLOP/s > Higher Is Better
HC ... 57.76  |===========
HBv2 . 93.79  |===================
HBv3 . 123.24 |========================
HBv4 . 323.36 |================================================================


HeFFTe - Highly Efficient FFT for Exascale 2.3
Test: r2c - Backend: FFTW - Precision: double - X Y Z: 256
GFLOP/s > Higher Is Better
HC ... 57.31  |==============
HBv2 . 91.92  |======================
HBv3 . 103.25 |=========================
HBv4 . 261.90 |================================================================


HeFFTe - Highly Efficient FFT for Exascale 2.3
Test: r2c - Backend: FFTW - Precision: double - X Y Z: 512
GFLOP/s > Higher Is Better
HC ... 60.88  |============
HBv2 . 91.48  |===================
HBv3 . 121.28 |=========================
HBv4 . 314.34 |================================================================


HeFFTe - Highly Efficient FFT for Exascale 2.3
Test: r2c - Backend: Stock - Precision: float - X Y Z: 512
GFLOP/s > Higher Is Better
HC ... 110.05 |============
HBv2 . 190.95 |====================
HBv3 . 232.17 |=========================
HBv4 . 596.23 |================================================================


HeFFTe - Highly Efficient FFT for Exascale 2.3
Test: c2c - Backend: Stock - Precision: double - X Y Z: 512
GFLOP/s > Higher Is Better
HC ... 31.57  |=============
HBv2 . 46.98  |===================
HBv3 . 56.22  |=======================
HBv4 . 154.65 |================================================================


HeFFTe - Highly Efficient FFT for Exascale 2.3
Test: r2c - Backend: Stock - Precision: double - X Y Z: 256
GFLOP/s > Higher Is Better
HC ... 60.57  |===============
HBv2 . 93.31  |=======================
HBv3 . 102.70 |=========================
HBv4 . 264.95 |================================================================


HeFFTe - Highly Efficient FFT for Exascale 2.3
Test: r2c - Backend: Stock - Precision: double - X Y Z: 512
GFLOP/s > Higher Is Better
HC ... 59.82  |============
HBv2 . 94.53  |===================
HBv3 . 117.73 |========================
HBv4 . 311.80 |================================================================


HeFFTe - Highly Efficient FFT for Exascale 2.3
Test: c2c - Backend: FFTW - Precision: float-long - X Y Z: 256
GFLOP/s > Higher Is Better
HC ... 58.55  |===============
HBv2 . 90.79  |=======================
HBv3 . 105.09 |==========================
HBv4 . 255.97 |================================================================


HeFFTe - Highly Efficient FFT for Exascale 2.3
Test: c2c - Backend: FFTW - Precision: float-long - X Y Z: 512
GFLOP/s > Higher Is Better
HC ... 62.90  |===========
HBv2 . 96.49  |=================
HBv3 . 135.95 |========================
HBv4 . 355.51 |================================================================


HeFFTe - Highly Efficient FFT for Exascale 2.3
Test: r2c - Backend: FFTW - Precision: float-long - X Y Z: 256
GFLOP/s > Higher Is Better
HC ... 122.77 |==================
HBv2 . 200.04 |==============================
HBv3 . 221.86 |=================================
HBv4 . 427.10 |================================================================


HeFFTe - Highly Efficient FFT for Exascale 2.3
Test: r2c - Backend: FFTW - Precision: float-long - X Y Z: 512
GFLOP/s > Higher Is Better
HC ... 113.94 |============
HBv2 . 191.14 |====================
HBv3 . 257.42 |==========================
HBv4 . 624.95 |================================================================


HeFFTe - Highly Efficient FFT for Exascale 2.3
Test: c2c - Backend: FFTW - Precision: double-long - X Y Z: 512
GFLOP/s > Higher Is Better
HC ... 33.55  |=============
HBv2 . 47.37  |===================
HBv3 . 57.23  |=======================
HBv4 . 159.26 |================================================================


HeFFTe - Highly Efficient FFT for Exascale 2.3
Test: c2c - Backend: Stock - Precision: float-long - X Y Z: 256
GFLOP/s > Higher Is Better
HC ... 59.55  |===============
HBv2 . 92.13  |========================
HBv3 . 105.36 |===========================
HBv4 . 247.73 |================================================================


HeFFTe - Highly Efficient FFT for Exascale 2.3
Test: c2c - Backend: Stock - Precision: float-long - X Y Z: 512
GFLOP/s > Higher Is Better
HC ... 57.92  |===========
HBv2 . 93.26  |==================
HBv3 . 124.60 |=========================
HBv4 . 323.70 |================================================================


HeFFTe - Highly Efficient FFT for Exascale 2.3
Test: r2c - Backend: FFTW - Precision: double-long - X Y Z: 256
GFLOP/s > Higher Is Better
HC ... 57.13  |=============
HBv2 . 88.61  |=====================
HBv3 . 106.63 |=========================
HBv4 . 273.12 |================================================================


HeFFTe - Highly Efficient FFT for Exascale 2.3
Test: r2c - Backend: FFTW - Precision: double-long - X Y Z: 512
GFLOP/s > Higher Is Better
HC ... 60.82  |============
HBv2 . 91.43  |===================
HBv3 . 120.96 |========================
HBv4 . 315.98 |================================================================


HeFFTe - Highly Efficient FFT for Exascale 2.3
Test: r2c - Backend: Stock - Precision: float-long - X Y Z: 512
GFLOP/s > Higher Is Better
HC ... 110.20 |============
HBv2 . 189.21 |====================
HBv3 . 233.80 |=========================
HBv4 . 590.93 |================================================================


HeFFTe - Highly Efficient FFT for Exascale 2.3
Test: c2c - Backend: Stock - Precision: double-long - X Y Z: 512
GFLOP/s > Higher Is Better
HC ... 31.58  |=============
HBv2 . 46.93  |===================
HBv3 . 56.27  |=======================
HBv4 . 154.57 |================================================================


HeFFTe - Highly Efficient FFT for Exascale 2.3
Test: r2c - Backend: Stock - Precision: double-long - X Y Z: 256
GFLOP/s > Higher Is Better
HC ... 60.89  |===============
HBv2 . 92.39  |=======================
HBv3 . 105.50 |==========================
HBv4 . 258.72 |================================================================


HeFFTe - Highly Efficient FFT for Exascale 2.3
Test: r2c - Backend: Stock - Precision: double-long - X Y Z: 512
GFLOP/s > Higher Is Better
HC ... 59.90  |============
HBv2 . 95.20  |====================
HBv3 . 118.24 |========================
HBv4 . 311.27 |================================================================


High Performance Conjugate Gradient 3.1
X Y Z: 104 104 104 - RT: 60
GFLOP/s > Higher Is Better
HC ... 26.00 |===================
HBv2 . 37.04 |===========================
HBv3 . 39.61 |=============================
HBv4 . 89.38 |=================================================================


High Performance Conjugate Gradient 3.1
X Y Z: 144 144 144 - RT: 60
GFLOP/s > Higher Is Better
HC ... 25.87 |===================
HBv2 . 36.09 |===========================
HBv3 . 38.97 |=============================
HBv4 . 88.52 |=================================================================


High Performance Conjugate Gradient 3.1
X Y Z: 160 160 160 - RT: 60
GFLOP/s > Higher Is Better
HC ... 25.56 |===================
HBv2 . 36.02 |===========================
HBv3 . 39.11 |=============================
HBv4 . 87.90 |=================================================================


Intel Open Image Denoise 2.0
Run: RT.hdr_alb_nrm.3840x2160 - Device: CPU-Only
Images / Sec > Higher Is Better
HC ... 1.85 |=======================================
HBv2 . 2.03 |===========================================
HBv3 . 1.72 |=====================================
HBv4 . 3.11 |==================================================================


Intel Open Image Denoise 2.0
Run: RT.ldr_alb_nrm.3840x2160 - Device: CPU-Only
Images / Sec > Higher Is Better
HC ... 1.85 |========================================
HBv2 . 2.01 |===========================================
HBv3 . 1.69 |====================================
HBv4 . 3.08 |==================================================================


Intel Open Image Denoise 2.0
Run: RTLightmap.hdr.4096x4096 - Device: CPU-Only
Images / Sec > Higher Is Better
HC ... 0.87 |============================================
HBv2 . 0.96 |================================================
HBv3 . 0.80 |========================================
HBv4 . 1.32 |==================================================================


Laghos 3.1
Test: Triple Point Problem
Major Kernels Total Rate > Higher Is Better
HC ... 156.52 |============================================
HBv2 . 183.82 |====================================================
HBv3 . 192.74 |======================================================
HBv4 . 228.15 |================================================================


Laghos 3.1
Test: Sedov Blast Wave, ube_922_hex.mesh
Major Kernels Total Rate > Higher Is Better
HC ... 247.49 |=======================================
HBv2 . 345.14 |=======================================================
HBv3 . 361.81 |=========================================================
HBv4 . 402.94 |================================================================


libxsmm 2-1.17-3645
M N K: 128
GFLOPS/s > Higher Is Better
HC ... 1284.8 |============
HBv2 . 1011.4 |==========
HBv3 . 2273.5 |======================
HBv4 . 6655.2 |================================================================


libxsmm 2-1.17-3645
M N K: 256
GFLOPS/s > Higher Is Better
HC ... 904.1  |========
HBv2 . 1128.3 |==========
HBv3 . 2045.7 |===================
HBv4 . 6908.6 |================================================================


libxsmm 2-1.17-3645
M N K: 32
GFLOPS/s > Higher Is Better
HC ... 384.9  |====
HBv2 . 164.8  |==
HBv3 . 1438.1 |===============
HBv4 . 6163.0 |================================================================


libxsmm 2-1.17-3645
M N K: 64
GFLOPS/s > Higher Is Better
HC ... 748.1  |========
HBv2 . 331.4  |====
HBv3 . 2413.7 |==========================
HBv4 . 5898.2 |================================================================


Liquid-DSP 1.6
Threads: 128 - Buffer Length: 256 - Filter Length: 57
samples/s > Higher Is Better
HC ... 1570633333 |=================
HBv2 . 4309133333 |================================================
HBv3 . 4216966667 |===============================================
HBv4 . 5412900000 |============================================================


Liquid-DSP 1.6
Threads: 176 - Buffer Length: 256 - Filter Length: 32
samples/s > Higher Is Better
HC ... 1536633333 |===============
HBv2 . 4275533333 |=========================================
HBv3 . 3864000000 |======================================
HBv4 . 6181766667 |============================================================


Liquid-DSP 1.6
Threads: 176 - Buffer Length: 256 - Filter Length: 57
samples/s > Higher Is Better
HC ... 1683033333 |==============
HBv2 . 4350100000 |=====================================
HBv3 . 4281533333 |====================================
HBv4 . 7095033333 |============================================================


Liquid-DSP 1.6
Threads: 176 - Buffer Length: 256 - Filter Length: 512
samples/s > Higher Is Better
HC ... 544626667  |===============
HBv2 . 924243333  |=========================
HBv3 . 814950000  |======================
HBv4 . 2221966667 |============================================================


NAMD 2.14
ATPase Simulation - 327,506 Atoms
days/ns < Lower Is Better
HC ... 0.52697 |===============================================================
HBv2 . 0.26505 |================================
HBv3 . 0.27111 |================================
HBv4 . 0.14380 |=================


NAS Parallel Benchmarks 3.4
Test / Class: BT.C
Total Mop/s > Higher Is Better
HC ... 106230.52 |=========
HBv2 . 241509.88 |====================
HBv3 . 313813.98 |==========================
HBv4 . 744413.90 |=============================================================


NAS Parallel Benchmarks 3.4
Test / Class: CG.C
Total Mop/s > Higher Is Better
HC ... 27619.05 |=======================
HBv2 . 36367.35 |==============================
HBv3 . 36681.43 |===============================
HBv4 . 74101.94 |==============================================================


NAS Parallel Benchmarks 3.4
Test / Class: FT.C
Total Mop/s > Higher Is Better
HC ... 55288.19  |===============
HBv2 . 98485.23  |==========================
HBv3 . 102122.36 |===========================
HBv4 . 230164.79 |=============================================================


NAS Parallel Benchmarks 3.4
Test / Class: IS.D
Total Mop/s > Higher Is Better
HC ... 1864.68  |=========
HBv2 . 3977.02  |===================
HBv3 . 5730.01  |===========================
HBv4 . 12967.37 |==============================================================


NAS Parallel Benchmarks 3.4
Test / Class: MG.C
Total Mop/s > Higher Is Better
HC ... 63404.01  |=========
HBv2 . 108985.72 |===============
HBv3 . 131635.41 |==================
HBv4 . 437417.16 |=============================================================


NAS Parallel Benchmarks 3.4
Test / Class: SP.C
Total Mop/s > Higher Is Better
HC ... 41543.94  |======
HBv2 . 104771.90 |===============
HBv3 . 205795.59 |=============================
HBv4 . 427298.99 |=============================================================


oneDNN 3.1
Harness: Recurrent Neural Network Training - Data Type: bf16bf16bf16 - Engine: CPU
ms < Lower Is Better
HC ... 707.32  |=================================
HBv2 . 1367.73 |===============================================================
HBv3 . 886.81  |=========================================
HBv4 . 533.49  |=========================


oneDNN 3.1
Harness: Recurrent Neural Network Inference - Data Type: bf16bf16bf16 - Engine: CPU
ms < Lower Is Better
HC ... 442.47 |===============================
HBv2 . 910.94 |================================================================
HBv3 . 529.97 |=====================================
HBv4 . 411.23 |=============================


OSPRay 2.12
Benchmark: particle_volume/ao/real_time
Items Per Second > Higher Is Better
HC ... 8.99618  |===============
HBv2 . 22.36680 |======================================
HBv3 . 24.47100 |=========================================
HBv4 . 36.65480 |==============================================================


OSPRay 2.12
Benchmark: particle_volume/scivis/real_time
Items Per Second > Higher Is Better
HC ... 8.87831  |===============
HBv2 . 22.17470 |======================================
HBv3 . 24.21970 |=========================================
HBv4 . 36.54460 |==============================================================


OSPRay 2.12
Benchmark: particle_volume/pathtracer/real_time
Items Per Second > Higher Is Better
HC ... 96.76  |==============================
HBv2 . 162.45 |==================================================
HBv3 . 167.50 |====================================================
HBv4 . 208.05 |================================================================


OSPRay 2.12
Benchmark: gravity_spheres_volume/dim_512/ao/real_time
Items Per Second > Higher Is Better
HC ... 9.52293  |================
HBv2 . 8.66888  |==============
HBv3 . 11.75010 |===================
HBv4 . 38.07690 |==============================================================


OSPRay 2.12
Benchmark: gravity_spheres_volume/dim_512/scivis/real_time
Items Per Second > Higher Is Better
HC ... 9.02689  |===============
HBv2 . 8.32323  |==============
HBv3 . 11.17230 |===================
HBv4 . 37.06240 |==============================================================


OSPRay 2.12
Benchmark: gravity_spheres_volume/dim_512/pathtracer/real_time
Items Per Second > Higher Is Better
HC ... 10.06 |====================
HBv2 . 13.94 |============================
HBv3 . 14.61 |=============================
HBv4 . 32.58 |=================================================================


Pennant 1.0.1
Test: sedovbig
Hydro Cycle Time - Seconds < Lower Is Better
HC ... 25.019560 |=============================================================
HBv2 . 5.915805  |==============
HBv3 . 6.277107  |===============
HBv4 . 3.581391  |=========


Pennant 1.0.1
Test: leblancbig
Hydro Cycle Time - Seconds < Lower Is Better
HC ... 10.645480 |=============================================================
HBv2 . 3.466885  |====================
HBv3 . 3.649317  |=====================
HBv4 . 2.122074  |============


PETSc 3.19
Test: Streams
MB/s > Higher Is Better
HC ... 151286.25 |===============
HBv2 . 197895.47 |====================
HBv3 . 284001.92 |=============================
HBv4 . 598417.70 |=============================================================


PostgreSQL 15
Scaling Factor: 1 - Clients: 500 - Mode: Read Only
TPS > Higher Is Better
HC ... 1353510 |===========================
HBv2 . 2467328 |=================================================
HBv3 . 2434749 |=================================================
HBv4 . 3161848 |===============================================================


PostgreSQL 15
Scaling Factor: 1 - Clients: 500 - Mode: Read Only - Average Latency
ms < Lower Is Better
HC ... 0.369 |=================================================================
HBv2 . 0.203 |====================================
HBv3 . 0.206 |====================================
HBv4 . 0.158 |============================


PostgreSQL 15
Scaling Factor: 1 - Clients: 800 - Mode: Read Only
TPS > Higher Is Better
HC ... 1159492 |=======================
HBv2 . 2481320 |==================================================
HBv3 . 2478917 |==================================================
HBv4 . 3146173 |===============================================================


PostgreSQL 15
Scaling Factor: 1 - Clients: 800 - Mode: Read Only - Average Latency
ms < Lower Is Better
HC ... 0.690 |=================================================================
HBv2 . 0.323 |==============================
HBv3 . 0.323 |==============================
HBv4 . 0.254 |========================


Timed Node.js Compilation 19.8.1
Time To Compile
Seconds < Lower Is Better
HC ... 330.61 |================================================================
HBv2 . 194.37 |======================================
HBv3 . 185.57 |====================================
HBv4 . 150.56 |=============================