AMD Expands Radeon AI PRO R9000 Series with New RDNA 4 Workstation GPUs

AMD has introduced two new additions to its Radeon AI PRO R9000 workstation graphics card lineup, the Radeon AI PRO R9700S and Radeon AI PRO R9600D. Both models are based on the advanced Navi 48 GPU, leveraging the latest RDNA 4 architecture to deliver enhanced performance for professional workloads.

Key Specifications and Performance

The Radeon AI PRO R9700S and R9600D are equipped with 32 GB of GDDR6 memory on a 256-bit interface, providing a peak memory bandwidth of 640 GB/s. This substantial memory capacity is designed to meet the demands of compute-intensive and visualization tasks, offering professionals a reliable hardware platform for large-scale projects.

The Radeon AI PRO R9700S features the full Navi 48 configuration, with 64 Compute Units (CUs) and 4,096 Stream Processors (SPs). It delivers up to 47.8 TeraFLOPS of FP32 performance at a boost clock of 2,920 MHz, with a total board power of 300 W. In contrast, the Radeon AI PRO R9600D is configured with 48 active CUs and 3,072 SPs, achieving approximately 24.8 TeraFLOPS of FP32 throughput at a 2,020 MHz boost clock, while maintaining a lower 150 W board power.

Cooling and Connectivity Features

Both the R9700S and R9600D utilize a passive heatsink design, moving away from the traditional onboard blower. This approach requires robust chassis airflow, typically provided by high-RPM industrial fans in server environments, to ensure optimal cooling and reliable operation. These workstation GPUs are tailored for deployment in server enclosures where noise is not a primary concern.

For power delivery, both cards adopt the new 12v-2x6 power connector standard. Display connectivity is supported through four DisplayPort 2.1a outputs, enabling advanced multi-display setups for professional visualization and compute workflows.

Availability

AMD has not yet announced pricing or shipping dates for the Radeon AI PRO R9700S and R9600D. These new workstation GPUs are positioned to provide professionals with scalable, high-performance solutions for demanding AI, compute, and visualization applications.