At GTC 2025, NVIDIA is unveiling the next generation of its professional graphics cards for desktops, laptops, and servers. Blackwel graphics adapters are designed for visualization and AI tasks.
The RTX 6000 PRO flagships feature 24,064 CUDA cores, 752 tensor cores, and 188 ray tracing cores in the GB202 chip. They are equipped with 96 GB of GDDR7 memory with ECC, operating on a 512-bit bus. This configuration means that NVIDIA used 3GB GDDR7 modules on both sides of the board.
In total, NVIDIA presented 12 professional graphics cards based on Blackwell architecture. These include a flagship model for workstations with two fans and TDP up to 600W, a turbine model for cluster workstations, and even a server version of the RTX PRO 6000.
Video cards for data centers
- NVIDIA RTX PRO 6000 Blackwell Server Edition
Video cards for desktop PCs
- NVIDIA RTX PRO 6000 Blackwell Workstation Edition
- NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Edition
- NVIDIA RTX PRO 5000 Blackwell
- NVIDIA RTX PRO 4500 Blackwell
- NVIDIA RTX PRO 4000 Blackwell
Video cards for laptops
- NVIDIA RTX PRO 5000 Blackwell
- NVIDIA RTX PRO 4000 Blackwell
- NVIDIA RTX PRO 3000 Blackwell
- NVIDIA RTX PRO 2000 Blackwell
- NVIDIA RTX PRO 1000 Blackwell
- NVIDIA RTX PRO 500 Blackwell
One of the RTX PRO 6000 cards has been labeled Max-Q. NVIDIA used to use this index to designate «maximally efficient» mobile graphics cards, but later abandoned it. Max-Q means that this card is optimized for energy efficiency. It has a TDP of 300 watts, which is half the power of the Workstation or Server version. NVIDIA is also launching a server graphics card that resembles the company’s earlier solutions. The model is fully passive cooling. It is intended for sale with third-party server configurations.
The RTX 6000 server and workstation versions deliver 125 TFLOPS of performance in single-precision computing. They are clocked at 2.6 GHz. The Max-Q card has 2.5 GHz. All three have 28 Gbps GDDR7 memory.
The RTX 5000 PRO is also equipped with a GB202 GPU (14080 CUDA cores). This card has 48GB of memory on a 384-bit memory bus and a TDP of 300W. RTX 4500 and RTX 4000 are based on the GB203 GPU. The former has 10496 CUDA cores and 32 GB of memory (256-bit) with 200W, while the latter has 8960 cores and 24 GB of 192-bit memory (140W). All new RTX PROs are equipped with DisplayPort 2.1b. Only the RTX 6000 version has a dual-flow cooler design, while the rest have a turbine design.
Source: VideoCardz
Spelling error report
The following text will be sent to our editors: