Empowering global eVTOL leaders to break through the bottleneck of simulation computing power and accelerate the air traffic revolution!
A certain civil aviation aircraft company in Shanghai focuses on the independent research and development of electric vertical takeoff and landing (eVTOL) aircraft, leading the innovation of the aviation industry with vertical takeoff and landing, low noise, high safety, and intelligent characteristics. The company is committed to building a green digital 3D transportation ecosystem, relying on advanced aviation technology and intelligent algorithms to provide efficient and environmentally friendly air travel solutions for global users.
<50db
Workstation full load noise as low as 50 decibels or less
Shorten the R&D cycle by 35%
Efficiency leap in fluid dynamics and structural analysis
Computational speed
Dual AMD EPYC platform+multi-core GPU cluster achieves doubled simulation efficiency
Project Background
A certain civil aviation aircraft company in Shanghai focuses on the independent research and development of electric vertical takeoff and landing (eVTOL) aircraft, leading the innovation of the aviation industry with vertical takeoff and landing, low noise, high safety, and intelligent characteristics. The company is committed to building a green digital 3D transportation ecosystem, relying on advanced aviation technology and intelligent algorithms to provide efficient and environmentally friendly air travel solutions for global users.
Challenge pain points
Insufficient computing power resources
CFD (Computational Fluid Dynamics) and FEA (Finite Element Analysis) simulations require processing TB level data, and traditional platform computing power cannot support real-time high-precision simulations
Software compatibility limitations
Advanced design tools in the industry require extremely high hardware parallel computing capabilities, making it difficult for existing systems to meet the concurrent demands of multitasking
Energy efficiency and noise balance
Balancing heat dissipation efficiency and noise control is necessary in high load scenarios to ensure adaptability to office environments
Solution
High performance heterogeneous computing platform
Deploy the Shuju Hongxin HW7325 workstation, equipped with dual AMD EPYC channels ™ The 9004 processor and NVIDIA A100 GPU cluster provide over 800 TFLOPS hybrid computing power and support CFD/FEA full process parallel computing.
Full stack software optimization
Deeply adapted to industrial software such as ANSYS and COMSOL, achieving dynamic scheduling of multi-threaded tasks and improving computing resource utilization by 90%
Intelligent wind liquid hybrid heat dissipation
Adopting a collaborative solution of air cooling and liquid cooling, it supports a 1500W TDP heat dissipation capacity, with a full load noise of ≤ 50 decibels, ensuring stable operation of the equipment 24/7.
Project Benefits
The speed of fluid dynamics simulation has been increased by 4 times, and the time required for a single simulation task has been reduced from hours to minutes, reducing research and development costs by 45%; Multi GPU cluster computing power supports real-time rendering of billion level grid models, increasing design iteration efficiency by 60% and shortening product launch cycles by 30%; The hybrid cooling solution achieves a 20% reduction in energy consumption, meets noise control standards, and significantly optimizes compatibility in office environments; Breaking through technological barriers has helped companies obtain three international patents in the eVTOL field, consolidating their leading position in industry technology.
Customer reviews
The Shuju Hongxin Workstation has provided revolutionary computing power support for our aircraft design, achieving dual breakthroughs in simulation efficiency and accuracy, and the project has advanced much faster than expected
