NVIDIA has unveiled two new professional graphics cards during SIGGRAPH 2025. These cards expand the Blackwell architecture lineup, introducing the RTX PRO 4000 SFF Edition and RTX PRO 2000. These compact GPUs are designed to accelerate AI for professional tasks while maintaining energy efficiency. Featuring fourth-generation RT Cores and fifth-generation Tensor Cores, these graphics cards deliver significant performance enhancements in smaller form factors.
The RTX PRO 4000 SFF Edition is equipped with 24 GB of ECC GDDR7 memory and offers 770 AI TOPS of throughput. This allows it to handle AI tasks 2.5 times faster and ray tracing 1.7 times better than previous models like the RTX A4000. Despite the updated specifications, NVIDIA has managed to maintain the same 70 W TGP as its predecessor.
On the other hand, the RTX PRO 2000 comes with 16 GB of ECC GDDR7 memory and targets mainstream design and AI workflows with 545 AI TOPS of power. Compared to the NVIDIA RTX A2000, it accelerates 3D modeling by 1.6 times, improves CAD performance by 1.4 times, and speeds up rendering. AI tasks such as image generation are 1.4 times quicker, and text generation is 2.3 times faster.
Both GPUs are ideal for compact workstations with limited space or power constraints. The RTX PRO 2000 will be available through PNY and TD SYNNEX, while both cards will be offered in complete systems from manufacturers like Dell, HP, and Lenovo. These graphics cards are expected to launch later this year, with pricing details yet to be disclosed.
RTX PRO 4000 Blackwell SFF Edition Specifications
- GPU architecture: NVIDIA Blackwell
- NVIDIA CUDA Cores: 8,960
- Tensor Cores: 5th Generation
- Ray Tracing Cores: 4th Generation
- GPU memory: 24 GB GDDR7 with ECC
- Memory interface: 192 bit
- Memory bandwidth: 432 GB/s
- System interface: PCIe 5.0 x8
- Display connectors: 4x Mini DisplayPort 2.1b
- Max simultaneous displays: 4x 3840x2160 @ 165 Hz / 2x 7680x4320 @ 100 Hz
- Video Engines: 2x NVENC (9th Gen), 2x NVDEC (6th Gen)
- Power consumption: Total board power: 70 W
- Thermal solution: Active
- Form factor: 2.7" x 6.6" L, dual slot, half height
- Graphics APIs: DirectX 12, Shader Model 6.7, OpenGL 4.6, Vulkan 1.4
- Compute APIs: CUDA 12.8, OpenCL 3.0
RTX PRO 2000 Blackwell Specifications
- GPU architecture: Blackwell
- NVIDIA CUDA Cores: 4,352
- Tensor Cores: 5th Generation
- Ray Tracing Cores: 4th Generation
- AI TOPS: 545
- Single-precision performance: 17 TFLOPS
- RT Core performance: 52 TFLOPS
- GPU memory: 16 GB GDDR7 with ECC
- Memory interface: 128-bit
- Memory bandwidth: 288 GB/s
- System interface: PCIe 5.0 x 8
- Display connectors: 4x mini DisplayPort 2.1b
- Max simultaneous displays: Up to four simultaneous displays
- Video Engines: 1x NVENC (9th Gen), 1x NVDEC (6th Gen)
- Power consumption: Max power consumption: 70 W
- Thermal solution: Active
- Form factor: 2.7" H x 6.6" L, dual slot
- Graphics APIs: DirectX 12, Shader Model 6.7, OpenGL 4.6, Vulkan 1.4
- Compute APIs: CUDA 12.8, OpenCL 3.0
