Mechanical analysis machine

A customer involved in mechanical engineering research and development asked us about a machine for using CAE software "RecurDyn" and "Matlab" for mechanism analysis (motion analysis).The conditions for machine selection are as follows.

・Budget is 100 million yen or less
・Emphasis on calculation speed

Based on the above conditions, we proposed the following configuration.

【Main Specifications】

CPU Core i9-12900K (3.20GHz 8-core + 2.40GHz 8-core)
memory 128GB
Storage 1 2TB M.2 SSD
Storage 2 4TB SSD S-ATA
Storage 3 4TB SSD S-ATA
video RTX-A2000
network on board (2.5GBase-T x1)
Housing + power supply Middle tower case + 1000W
OS Microsoft Windows 11 Pro 64 bit

There are two things to keep in mind when considering a machine for RecurDyn.

■ Points
・RecurDyn may have no effect on calculation speed even with too many CPU cores.
・RecurDyn alone does not perform GPU acceleration.

Reference: RecurDyn ServicePack (FunctionBay) *Jumps to an external site

From the above points,

・Is the number of cores effective in Matlab?
・Are you using GPU acceleration?

Since the appropriate configuration varies depending on the customer, we asked the customer to check the usage situation on the Matlab side.

■ Questions and requests from customers
・We are using a tool called SMP (parallel processing computation), but is it still effective to increase the number of cores?
Reference: SMP parallel processing calculation *Jumps to an external site

The SMP manufacturer page states that it supports 8-core and 16-core parallel processing calculations (SMP), so paradoxically we decided that it would be better to set it to a maximum of 16 cores.Therefore, the CPU adopts "Core i8-8K" with high performance core 9 core + high efficiency core 12900 core specification. 8-core calculations are handled by high-performance cores, and 16-core calculations are handled by using high-performance cores + high-efficiency cores.

・When using MATLAB, I don't know how to use it considering the number of cores ・I don't know how to write code, but I want to analyze big data in the future.At that time, I want to make the calculation speed as fast as possible.

The current situation is that CPU parallelization is not used much in standard programs such as "R" and "SPSS", which are often used for analysis purposes.You can use parallelism, but it's not the type that is more efficient with more cores.Therefore, we believe that there is no problem in using the core specifications of this example.
On the other hand, if you plan to analyze big data in the future, we recommend upgrading the storage capacity and speed.

In addition, the A2000 was selected as the video card because the customer told us that they would also use CAD.

The configuration of this case study is based on the conditions given by the customer.
Please feel free to contact us even if you are considering different conditions from what is posted.

 

■FAQ

・What is RecurDyn?
RecurDyn is CAE software for mechanical analysis (motion analysis) to evaluate the motion mechanism of mechanical systems.Originally used for tank track development, etc., it was commercialized for the private sector and was used by construction heavy equipment manufacturers, and later, by supporting the analysis of precision machinery, it expanded into the fields of automobiles and robots. also spread.

Reference: RecurDyn *Jumps to an external site

・What are “high-performance cores” and “high-efficiency cores”?
The type of core installed in Alder Lake, the Intel 12th generation desktop CPU.
The high performance core (Golden Cove Performance Core: P core) emphasizes processing performance, and the high efficiency core (Gracemont Efficiency Core: E core) emphasizes power efficiency.

In the case of Core i9-12900K, the specifications of each core are as follows.

High performance core (P core): 3.2GHz to 5.2GHz 8 cores 16 threads
High efficiency core (E core): 2.4GHz to 3.9GHz 8 cores 8 threads