NVIDIA RTX PRO 6000 “Blackwell” Series Unveiled: Top-Tier GB202 GPU Featuring 24,000 Cores, 96 GB VRAM, and Up to 600W TDP

NVIDIA has recently unveiled its groundbreaking RTX PRO 6000 “Blackwell”series of GPUs, specially crafted for the prosumer and data center markets, offering exceptional computational power.

The RTX PRO 6000 “Blackwell”: A New Era for Graphics Processing

Today marks the official release of NVIDIA’s latest advancement in graphics technology with the RTX PRO 6000 “Blackwell”GPU. This state-of-the-art graphics card surpasses even the capabilities of the gaming-oriented RTX 5090, focusing on robust performance for prosumers and enterprises. The lineup includes three distinct variants: a standard design, a Max-Q model, and a server-optimized version tailored to specific consumer needs.

Equipped with an impressive 24, 064 GPU cores, the “Blackwell”series outperforms the RTX 5090, which has 21, 760 cores, representing a 10.5% increase. In addition, users will benefit from 752 tensor cores and 188 RT cores. This architecture enables the card to deliver up to 125 TFLOPs of FP32 performance and a remarkable 4000 AI TOPS, showcasing a significant boost in computational efficiency. However, the standout feature is its unprecedented memory capability.

In contrast to the NVIDIA GeForce RTX 5090’s 32 GB GDDR7 memory, the RTX PRO 6000 “Blackwell”is equipped with an astonishing total of 96 GB of GDDR7 (ECC) memory, utilizing a wide 512-bit bus. With data transfer speeds reaching 28 Gbps, the card achieves a total bandwidth of up to 1.8 TB/s.

Regarding total board power (TBP), the NVIDIA RTX PRO 6000 is rated at 600W, utilizing a single 12V-2x6 16-pin power interface. Cooling solutions are imperative for managing the thermal output of such powerful hardware. To that end, NVIDIA’s thermal engineering team has devised a robust dual-fan and dual-slot cooling system for the standard model.

Each variant of the RTX PRO 6000 addresses specific market needs; for example, the Max-Q model is designed with a blower-style configuration aimed at workstation users, while the server variant incorporates a passive cooling system for optimal performance in data centers. The Max-Q edition operates with a thermodynamic design power (TDP) of just 300W.

As for the pricing, prospective buyers should prepare to invest significantly, with estimates for the workstation flagship hovering between $10, 000 and $15, 000. Canadian retailer Direct Dial lists the card for approximately CAD 11, 933, which translates to about $8, 300 USD. This substantial investment reflects the card’s capabilities, as the RTX 5090 is currently priced around $3, 000 USD, making the RTX PRO 6000 “Blackwell”over two times more costly.

NVIDIA Workstation Graphics Card Overview

For more details on the RTX PRO 6000 “Blackwell, ”check out this source and images.

Related Articles:

NWinfo Version 1.4.3 Release Details

3:31September 4, 2025

ASUS Introduces TUF Gaming Series 5: New 27 and 34-Inch Monitors with Refresh Rates Up to 300Hz

20:55September 3, 2025

System Integrator Reveals Ryzen 7 9700F Features 5.4 GHz Boost Clock; Retains 9700X Configuration

20:04September 3, 2025

Hugging Face CEO Warns U.S. AI Industry Will Suffer If Closed-Source Adoption Continues

19:53September 3, 2025