JAX Machine Learning and Abaqus Analysis Workstation

A customer involved in mechanical engineering research inquired about a PC configuration that would primarily be used for machine learning and numerical simulations using Google ML's JAX, but also for analysis using general-purpose finite element method software (such as Abaqus).
The client requested a Blackwell generation GPU and an Intel CPU, with as much GPU memory as possible.
The budget is around 100 million yen.

CPU	Intel Core Ultra 9 285K 3.70GHz(8C/8T) + 3.20GHz(16C/16T)
memory	Total 64GB DDR5 6400 32GB x 2
Storage 1	1TB SSD M.2 NVMe Gen4
Video	NVIDIA RTX PRO4500 32GB
network	on board(2.5GBase-T x1) Wi-Fi, Bluetooth
Housing + power supply	Mid-tower chassis + 1000W 80PLUS PLATINUM
OS	Microsoft Windows 11 Professional 64bit

GPU selection

In machine learning with JAX,GPU memory capacity and computing performanceIs important.
The Blackwell generation GPU is suitable for machine learning, but considering your budget, we recommend the RTX PRO 4500 (32GB VRAM).
The GeForce RTX 5090, which has the same video memory capacity and high processing power, is also a candidate for consideration, but it requires improved power supply and cooling performance.The best option within your budgetWe adopted the RTX PRO 4500 as the graphics card.

Regarding CPU and memory selection

The processing speed of the finite element analysis software Abaqus is directly linked to the single-thread performance of the CPU.
Therefore, this configuration uses the Intel Core Ultra 9 285K, which has high single-thread performance, to achieve both shorter analysis times and stability.
The memory is DDR5-6400, with two 32GB cards (64GB total), and can be expanded up to 128GB in the future using the available slot.

Power capacity and stable operation

A 1000W power supply is used to support simultaneous operation of a high-performance GPU and CPU.
By providing a power margin, shutdowns and performance degradation due to load fluctuations at peak times can be prevented.Ensure system stabilitydoing.

For those who are active in these fields Machine learning Structural analysis Material engineering Computational Mechanics Numerical analysis

Tegara's custom-made PC production service not only caters to initial use, but also supports system expansion in anticipation of future expansion of research scale.
We not only propose configurations that meet various software requirements, but also accept consultations regarding the construction of an entire research environment.
Please feel free to contact us and we will provide the best solution to suit your needs.

Keyword What is JAX? JAX is a numerical calculation library for Python developed by Google, which enables high-speed parallel processing through automatic differentiation and JIT compilation. It is primarily used in the fields of machine learning and scientific computing, and is particularly effective in deep learning and large-scale simulations. It can easily accelerate NumPy code with GPU/TPU support, and also features flexible function conversion functions (vmap, scan, etc.). reference:JAX Documentation *Jumps to an external site What is Google ML? Google-ML is a collective term for machine learning (ML) technologies and tools provided by Google, a group of software that learns patterns from large amounts of data and enables predictions and content generation. It is widely used for experiments and advanced analysis using data in a variety of formats, including images, text, audio, and video. It can be flexibly used in on-premise and cloud environments, and is highly rated for its ability to efficiently build and operate ML models without specialized knowledge. reference:Google Cloud Machine Learning *Jumps to an external site What is Abaqus? Abaqus is an advanced finite element analysis (FEA) software provided by Dassault Systèmes, with strengths in nonlinear and multiphysics analysis. It is widely used in design and development sites in various industries, including aerospace, automobiles, defense, and home appliances, as well as in university research institutes, and is suitable for simulating complex physical phenomena. Its major features are its user-friendly GUI and script-based automation capabilities, as well as its ability to handle detailed material models and complex analyses. reference:Dassault Systèmes SIMULIA – Abaqus *Jumps to an external site

■ Click here for details and inquiries about this PC case JAX Machine Learning and Abaqus Analysis Workstation * Please enter the name of the case or your desired conditions.