AMD launches powerful silent graphics card for ai workloads

AMD is targeting the growing demand for artificial intelligence infrastructure with its new Radeon AI Pro R9700S. This workstation graphics card is unique because it offers high end performance without the noise of traditional cooling fans. By using a passive cooling system, the card relies on the existing airflow within a server rack to stay cool. This design is specifically intended for data centers where multiple GPUs are packed closely together, as removing individual fans can improve reliability and reduce the overall noise in the facility.

The specifications for the R9700S are impressive for a silent card. It features 32GB of GDDR6 memory, which is essential for handling large AI models and complex datasets. The card is built on AMD’s RDNA 4 architecture and delivers a peak compute power of 47.8 TFLOPS. Despite having no fans, it still has a significant power draw of 300W. This means that while the card itself is silent, the server it lives in will need powerful external fans to move enough air across the large heatsink to prevent overheating during intense processing tasks.

 

 

In addition to the R9700S, AMD also revealed a slightly smaller version called the R9600D. Both cards are designed to be fully compatible with Linux and AMD’s professional software tools, making them easy to integrate into existing AI training and inference setups. These cards are geared toward businesses that want to offload heavy calculations from their main processors to dedicated hardware. By offering a silent, high capacity option, AMD is giving enterprise customers a way to build denser, more efficient AI servers.

 

 

You can expect to see the Radeon AI Pro R9700S appearing in pre-built workstations and server configurations starting early this year. Several manufacturers have already confirmed they will be using these new cards in their professional hardware lineups. While these are not meant for home gaming PCs, the technology inside shows how AMD is pushing to compete with Nvidia in the professional AI market. If you are looking to upgrade your company’s compute power, you should check with enterprise hardware suppliers for availability and pricing details.