Welcome Here  Shenzhen Mingjiada Electronics Co., Ltd.

Recently, NVIDIA officially announced the launch of the NVIDIA RTX PRO server before Computex2025 and emphasized that IT is designed for enterprise and industrial AI computing, aiming to accelerate the transformation of trillion-dollar enterprise IT data center infrastructure into AI factories.

The NVIDIA RTX PRO server is a high-performance server specifically designed for enterprise and industrial AI computing, aiming to accelerate complex workloads such as AI inference, graphics rendering, and scientific computing. The core of this server is equipped with eight RTX 6000 Blackwell professional graphics cards. Each graphics card is equipped with 24,064 CUDA cores, 96GB GDDR7 video memory and 512-bit memory bus. The video memory bandwidth is as high as 13TB/s, and the overall video memory capacity reaches 800GB.

The RTX PRO server supports an omnidirectional bandwidth of 800GB/s through the ConnectX-8 SuperNIC interface. Moreover, the power consumption of a single card is 600W, and the power consumption of the entire machine exceeds 5000W. At FP32 precision, its AI performance can reach 30 PFLOPS, and its graphics rendering performance is 3PFlops, enabling it to efficiently run various high-performance AI models.

Furthermore, this server supports up to eight RTX PRO 6000 Blackwell Gpus and integrates the NVIDIA BlueField-3 data processing Unit (DPU) as well as the NVIDIA ConnectX®-8 SuperNIC with an integrated PCIe Gen 6 switch. It can become the core driving force of an AI factory.

The goal of the RTX PRO server is to promote the transformation of enterprise IT data centers into AI factories and provide powerful computing support for designers, developers, engineers and scientists. It is not only applicable to the GPU fusion acceleration in data centers, but also can achieve high-performance AI tasks anytime and anywhere through remote technology.

The NVIDIA RTX PRO server, with its outstanding performance, flexible scalability and powerful AI computing capabilities, is becoming an indispensable tool in enterprises and industrial fields, helping users achieve breakthroughs in areas such as AI, graphics rendering and scientific computing.

It is worth mentioning that in his opening speech at Computex2025, NVIDIA CEO Jensen Huang announced that NVIDIA will join hands with TSMC and Foxconn to build an AI supercomputer in Taiwan, China. He said, "We are on the verge of a prosperous future. The value of the chip industry has reached 300 billion US dollars, and the opportunities in data centers are transforming into a market worth nearly one trillion US dollars, all driven by AI factories and infrastructure."

Website: www.hkmjd.com