readmd.md for benchmark is updated

This commit is contained in:
Hamidreza 2025-04-28 12:25:53 +03:30
parent cd051a6497
commit 8986c47b69
4 changed files with 65 additions and 45 deletions

Binary file not shown.

After

Width:  |  Height:  |  Size: 129 KiB

View File

@ -1,76 +1,96 @@
# Benchmarks of rotating drum (v-1.0)
## Problem definition
**The rotatingDrum case was calculated using well-known discrete element commercial software and phasicFlow, and the calculation time and memory usage were tested when the number of particles was 250 k, 500 k, 1 m, 2 m, 4 m and 8 m.**
# Rotating Drum Benchmark (phasicFlow v-1.0)
## Overview
This benchmark compares the performance of phasicFlow with a well-stablished commercial DEM software for simulating a rotating drum with varying particle counts (250k to 8M particles). The benchmark measures both computational efficiency and memory usage across different hardware configurations.
## Simulation Setup
<div align="center">
<img src="./images/commerical DEM snapshot.png" alt="description" width="400" />
<img src="./images/commerical DEM snapshot.png" style="width: 400px;" />
<div align="center">
<p>Figure 1. Commercial DEM simulation snapshot</p>
</div>
</div>
<div align="center">Figure 1. The snapshot of commerical DEM simulation</div>
<div align="center">
<img src="./images/phasicFlow snapshot.png" alt="description" height="600" />
<img src="./images/phasicFlow snapshot.png" style="width: 400px;" />
<div align="center">
<p>Figure 2. phasicFlow simulation snapshot</p>
</div>
</div>
### Hardware Specifications
<div align="center">Figure 2. The snapshot of phasicFlow simulation</div>
<div align="center">
Table 1. Hardware specifications used for benchmarking.
</div>
**The performance tests were conducted on laptop computers and workstation computers respectively. The configuration information of the laptop computers and workstation computers is shown in Table 1.**
<div align="center">Table 1. The parameters of laptop and workstation.</div>
| computer | CPU | GPU | Operating system |
| System | CPU | GPU | Operating System |
| :---------: | :----------------------: | :--------------------------: | :--------------: |
| Laptop | Intel i9-13900HX 2.2 GHz | NVIDIA Geforce RTX 4050Ti 6G | Windows 11 24H2 |
| workstation | Intel xeon 4210 2.2 GHz | NVIDIA RTX A4000 16G | Ubuntu 22.04 |
| Laptop | Intel i9-13900HX 2.2 GHz | NVIDIA GeForce RTX 4050Ti 6G | Windows 11 24H2 |
| Workstation | Intel Xeon 4210 2.2 GHz | NVIDIA RTX A4000 16G | Ubuntu 22.04 |
**A total of 6 groups of working condition tests were carried out, with the particle numbers being 250 k, 500 k, 1 m, 2 m, 4 m and 8 m respectively. The particle diameter, drum length and drum diameter under each working condition are shown in Table 2.**
### Simulation Parameters
<div align="center">Table 2. The parameters setting of different rotating drum.</div>
<div align="center">
Table 2. Parameters for rotating drum simulations.
</div>
| Case name | Particle diameter | Particle numbers | Rotating drum length | Rotating drum radius |
| Case | Particle Diameter | Particle Count | Drum Length | Drum Radius |
| :-------: | :---------------: | :--------------: | :------------------: | :------------------: |
| 250 k | 6 mm | 250,000 | 0.8 m | 0.2 m |
| 500 k | 5 mm | 500,000 | 0.8 m | 0.2 m |
| 1 m | 4 mm | 1,000,000 | 0.8 m | 0.2 m |
| 2 m | 3 mm | 2,000,000 | 1.2 m | 0.2 m |
| 4 m | 3 mm | 4,000,000 | 1.6 m | 0.2 m |
| 8 m | 2 mm | 8,000,000 | 1.6 m | 0.2 m |
| 250k | 6 mm | 250,000 | 0.8 m | 0.2 m |
| 500k | 5 mm | 500,000 | 0.8 m | 0.2 m |
| 1M | 4 mm | 1,000,000 | 0.8 m | 0.2 m |
| 2M | 3 mm | 2,000,000 | 1.2 m | 0.2 m |
| 4M | 3 mm | 4,000,000 | 1.6 m | 0.2 m |
| 8M | 2 mm | 8,000,000 | 1.6 m | 0.2 m |
**Table 3 and Figure 3 show the calculation time of phasicFlow and well-known discrete element commercial software under different working conditions on a laptop. It can be seen from the table that the increase in calculation time is linearly proportional to the increase in the number of particles. The calculation speed of phasicFlow is about 20% faster than that of well-known discrete element commercial software. The calculation speed of using A4000 graphics card is about 30% faster than that of 4050Ti.**
The time step for all simulations was set to 1.0e-5 seconds and the simulation ran for 4 seconds.
<div align="center">Table 3 The calculation time of different software.</div>
## Performance Comparison
| Software | 250 k | 500 k | 1 m | 2 m | 4 m | 8 m |
### Execution Time
<div align="center">
Table 3. Total calculation time (minutes) for different configurations.
</div>
| Software | 250k | 500k | 1M | 2M | 4M | 8M |
| :---------------: | :----: | :-----: | :-----: | :-----: | :-----: | :------: |
| phasicFlow-4050Ti | 54 min | 111 min | 216 min | 432 min | | |
| commerical DEM-4050Ti | 68 min | 136 min | 275 min | 570 min | | |
| phasicFlow-4050Ti | 54 min | 111 min | 216 min | 432 min | - | - |
| Commercial DEM-4050Ti | 68 min | 136 min | 275 min | 570 min | - | - |
| phasicFlow-A4000 | 38 min | 73 min | 146 min | 293 min | 589 min | 1188 min |
The execution time scales linearly with particle count. phasicFlow demonstrates approximately:
- 20% faster calculation than the well-established commercial DEM software on the same hardware
- 30% performance improvement when using the NVIDIA RTX A4000 compared to the RTX 4050Ti
<div align="center">
<img src="./images/Calculation time under different particle numbers using commerical DEM and phasicFlow.png" alt="description" height="600" />
<img src="./images/performance1.png" style="width: 500px;" />
<p>Figure 3. Calculation time comparison between phasicFlow and the well-established commercial DEM software.</p>
</div>
<div align="center">Figure 3 The calculation time of different software.</div>
### Memory Usage
<div align="center">
Table 4. Memory consumption for different configurations.
</div>
**Table 4 and Figure 4 show the memory usage of phasicFlow and well-known discrete element commercial software on a laptop under different working conditions. It can be seen from the table that phasicFlow uses about 0.7 GB of memory per million particles, while the well-known discrete element commercial software uses about 1.2 GB of memory per million particles.**
<div align="center">Table 4 The memory usage of different software.</div>
| Software | 250 k | 500 k | 1 m | 2 m | 4 m | 8 m |
| Software | 250k | 500k | 1M | 2M | 4M | 8M |
| :---------------: | :-----: | :-----: | :-----: | :-----: | :-----: | :-----: |
| phasicFlow-4050Ti | 252 MB | 412 MB | 710 MB | 1292 MB | | |
| commerical DEM-4050Ti | 485 MB | 897 MB | 1525 MB | 2724 MB | | |
| phasicFlow-4050Ti | 252 MB | 412 MB | 710 MB | 1292 MB | - | - |
| Commercial DEM-4050Ti | 485 MB | 897 MB | 1525 MB | 2724 MB | - | - |
| phasicFlow-A4000 | 344 MB | 480 MB | 802 MB | 1386 MB | 2590 MB | 4966 MB |
Memory efficiency comparison:
- phasicFlow uses approximately 0.7 GB of memory per million particles
- Commercial DEM software uses approximately 1.2 GB of memory per million particles
- phasicFlow shows ~42% lower memory consumption compared to the commercial alternative
- The memory usage scales linearly with particle count in both software packages. But due to memory limitations on GPUs, it is possible to run larger simulation on GPUs with phasicFlow.
<img src="./images/Memory usage under different particle numbers using commerical DEM and phasicFlow.png" alt="description" height="600" />
## Run Your Own Benchmarks
<div align="center">Figure 4 The memory usage of different software.</div>
The simulation case setup files are available in this folder for users interested in performing similar benchmarks on their own hardware. These files can be used to reproduce the tests and compare performance across different systems.