AMD has formally launched the Radeon Graphics Cards are RX 9000 Series: Radeon RX 9060 XT and AMD Radeon AI PRO R9700, powerful graphics cards designed for sophisticated gaming applications, professional visualization, and AI workloads. This GPU, which was introduced at COMPUTEX 2025, is made to handle the demands of contemporary AI and gaming tasks.
AMD Radeon RX 9060 XT GPU
| Model | Compute Units | VRAM | Game Clock (GHz) | Boost Clock1 (GHz) | Memory Interface | Infinity Cache | TBP | Price (USD SEP) | |
| AMD Radeon RX 9060 XT 8GB | 32 | 8 GB | 2.53 | Up to 3.13 | 128-bit | 32 MB | Starting at 150W | $299 | |
| AMD Radeon RX 9060 XT 16GB | 32 | 16 GB | 2.53 | Up to 3.13 | 128-bit | 32 MB | Starting at 160W | $349 | |
Pricing And availability
Later this year, top board partners Acer, ASRock, ASUS, Gigabyte, PowerColor, Sapphire, Vastarmor, XFX, and Yeston will sell AMD Radeon RX 9060 XT graphics cards. AMD Radeon RX 9060 XT 8GB and 16GB SEPs are $299 and $349, respectively.
AMD’s most recent Radeon RX 9000 Series places a strong emphasis on superior graphics capabilities and AI acceleration.
Also Read About AMD Radeon RX 9070 XT vs RX 7900 XT: Which GPU Is Better?
These features establish the R9700 as a potent instrument capable of managing intricate calculations and sizable AI models, which makes it appropriate for applications like machine learning inference and training.
AMD Radeon AI PRO R9700
Architectural Features
AMD’s RDNA 4 architecture, upon which the Radeon AI PRO R9700 is based, provides notable enhancements over RDNA 3. Notable improvements consist of:
- Enhanced AI computing: Thanks to two AI accelerators per computing unit, there is more than four times the AI compute capacity of RDNA 3.
- Support for New Data Types: Increased adaptability in managing diverse AI tasks.
- Support for Sparsity: Effective handling of sparse data structures, which are frequently used in AI applications.
- The R9700 can now meet the demands of contemporary AI workloads with up to 1,557 TOPS of performance thanks to these enhancements.
Support for Ecosystems and Software
The Radeon AI PRO R9700’s smooth integration with well-known AI frameworks and tools has been guaranteed by AMD:
- Framework Compatibility: Microsoft DirectML and other popular machine learning frameworks are supported.
- AMD’s ROCm open software platform works with Radeon RX graphics cards like the R9700, making machine learning development easier.
- Amuse 2.0 for AMD Radeon graphics cards simplifies generative AI model exploration.
Developers and researchers may efficiently use the R9700 for a variety of AI applications thanks to these software connections.
Accessibility
It is anticipated that the AMD Radeon AI PRO R9700 will go on sale in July 2025. AMD’s dedication to offering high-performance solutions for professional workloads and AI is demonstrated by its launch.
AMD Radeon AI PRO R9700: Designed for Workstations with AI
The AMD Radeon AI PRO R9700 is a reliable tool for professionals working to further AI development since it speeds up data-intensive activities like local inference and model finetuning. The R9700 is designed for high-performance AI with on-device control and flexibility, with second-generation AI accelerators that can give up to twice the throughput of the previous generation2 and complete support for the ROCm software stack on Linux, with support for ROCm on Windows on the horizon.
Additionally, the Radeon AI PRO R9700 grows well in multi-GPU setups, increasing memory and processing power for complex models and concurrent jobs. Because of its adaptability, it is perfect for sophisticated workstations handling demanding simulations, real-time rendering, or several AI workloads at once.
| Model | Compute Units | VRAM | Game Clock (GHz) | Boost Clock1 (GHz) | Memory Interface | Infinity Cache | TBP |
| AMD Radeon AI PRO R9700 | 64 | 32 GB | 2.35 | Up to 2.92 | 256-bit | 64 MB | 300W |
AMD Radeon AI PRO R9700 Benchmark
Benchmark Performance (Estimated, based on early data):
| Benchmark/Test | Performance | Comparison |
|---|---|---|
| Blender (Render) | ~18% faster than RTX 5000 Ada | Superior for heavy visual workloads |
| AI Inference (INT4 Sparse) | Up to 1,531 TOPS | Up to 4x faster than RDNA 3 GPUs |
| Stable Diffusion (v1.5) | Generates images 2x faster than RX 7900 XT | Optimized for text-to-image tasks |
| 3DMark Time Spy Extreme | ~20,500 (Estimated) | Close to RTX 4080-level performance |
| Gaming (4K, Ultra) | 60–90 FPS in AAA titles w/ FSR 4 | Comparable to RTX 4070 Ti/4080 |
| Ray Tracing (4K) | Improved vs RDNA 3 but behind Ada Lovelace | Good for hybrid pro/gaming workloads |
| AI Workload (TensorFlow) | 2x throughput vs Radeon PRO W7800 | Suitable for large model training/inference |
AMD Radeon AI PRO R9700 Specs
| Category | Specification |
|---|---|
| GPU Architecture | RDNA 4 |
| GPU Chip | Navi 48 |
| Compute Units (CUs) | 64 |
| AI Accelerators | 128 (2 per CU) |
| Stream Processors | 4,096 |
| Ray Accelerators | Next-gen RDNA 4-based |
| Memory | 32GB GDDR6 |
| Memory Speed | 20 Gbps |
| Memory Interface | 256-bit |
| Memory Bandwidth | 640 GB/s |
| Peak FP16 Performance | 96 TFLOPS |
| INT4 Sparse Performance | 1,531 TOPS |
| TBP (Total Board Power) | 300W |
| Interface | PCIe 5.0 x16 |
| Display Outputs | 4 × DisplayPort 2.1a |
| AI Framework Support | Microsoft DirectML, ROCm |
| Driver Support | Radeon PRO Software, Enterprise-certified |
| Launch Date | Expected July 2025 |
| Target Audience | AI developers, researchers, pro visualization |
Final Thoughts
The AMD Radeon AI PRO R9700, which combines strong AI acceleration, cutting-edge gaming features, and professional-grade performance, is a major advancement in GPU technology. It is based on the new RDNA 4 architecture and offers 32GB of fast GDDR6 memory, up to 1,531 TOPS of AI computation, and optimized support for graphics-intensive applications and AI frameworks.
Also Read About AMD Radeon RX 6750 XT VS RX 7800 XT: which GPU is better?
The R9700 is a unique GPU for developers and producers due to its capabilities for large-scale AI tasks, compatibility with AMD’s ROCm platform, and improved gaming through AI-powered FSR 4.