PNY NVIDIA RTX PRO 6000 Blackwell Workstation Edition 96GB Professional Graphics Card

The PNY NVIDIA RTX PRO 6000 Blackwell Workstation Edition is a flagship professional graphics card built on NVIDIA Blackwell architecture for demanding AI, engineering, scientific computing, visualization, simulation, rendering and content-creation workloads. It features 24,064 CUDA cores, 752 Tensor Cores, 188 RT Cores and 96GB of ECC-enabled GDDR7 memory with 1,792 GB/s of memory bandwidth.

The card uses a full-length, extra-high-profile dual-slot design with a double-flow-through thermal solution and a maximum power consumption of 600W. It provides four DisplayPort 2.1 outputs and supports NVIDIA RTX PRO Sync, making it suitable for high-end professional workstation environments.

Key Features

• NVIDIA Blackwell architecture with 24,064 CUDA cores
• 96GB GDDR7 memory with ECC and 1,792 GB/s bandwidth
• 752 fifth-generation Tensor Cores and 188 fourth-generation RT Cores
• 600W maximum power consumption with PCIe CEM5 16-pin power connector
• Four DisplayPort 2.1 outputs with NVIDIA RTX PRO Sync compatibility

Complete Technical Specifications

• Brand: PNY
• Model: NVIDIA RTX PRO 6000 Blackwell Workstation Edition
• Series: NVIDIA RTX PRO Blackwell
• GPU/CPU/Product Type: Professional Workstation Graphics Card
• Architecture: NVIDIA Blackwell
• Memory: 96GB
• Memory Type: GDDR7 with ECC
• Memory Bus: 512-bit
• Memory Speed: Not Available
• Clock Speed: Not Available
• Interface: PCI Express 5.0 x16
• Dimensions: 5.4" H × 12" L
• Power Consumption: 600W maximum
• Power Connector: 1 × PCIe CEM5 16-pin
• Recommended PSU: Not Available
• Cooling: Double Flow Through
• Operating System: Windows 11 64-bit, Windows 10 64-bit, Red Hat Enterprise Linux 7.x, SUSE Linux Enterprise Desktop 15.x, Fedora 31, Ubuntu 18.04, FreeBSD 11.x, Solaris 11
• CUDA Cores: 24,064
• Tensor Cores: 752
• RT Cores: 188
• Single-Precision Performance: 125 TFLOPS
• AI Performance: 4,000 AI TOPS
• RT Core Performance: 380 TFLOPS
• Memory Bandwidth: 1,792 GB/s
• Display Connectors: 4 × DisplayPort 2.1
• Maximum Displays: 4
• Maximum Digital Display Resolution: 7680 × 4320
• Form Factor: XHFL Dual Slot
• Foundry: TSMC
• Process Size: 4N NVIDIA Custom Process
• Transistors: 92.2 billion
• Die Size: 750 mm²
• Multi-Instance GPU: Up to 4 × 24GB, 2 × 48GB, or 1 × 96GB
• NVENC: 4 ×
• NVDEC: 4 ×
• JPEG Engines: 4 ×
• Frame Lock: Compatible with NVIDIA RTX PRO Sync
• NVIDIA 3D Vision: Supported via 3-pin Mini DIN
• PNY Part Number: VCNRTXPRO6000B-PB

GPU Architecture

The RTX PRO 6000 Blackwell Workstation Edition uses NVIDIA Blackwell architecture with 24,064 CUDA cores, 752 Tensor Cores and 188 RT Cores. PNY specifies peak performance of 125 TFLOPS single precision, 380 TFLOPS RT Core performance and up to 4,000 AI TOPS.

Memory

The card is equipped with 96GB of GDDR7 memory with ECC, using a 512-bit memory interface and delivering 1,792 GB/s of memory bandwidth. This large ECC-enabled memory pool is designed for professional workloads involving large datasets, complex scenes, AI models and high-resolution assets.

Display Support

Four DisplayPort 2.1 outputs are provided. PNY specifies support for up to four displays, with configurations including four 4096 × 2160 displays at 120Hz, four 5120 × 2880 displays at 60Hz, or two 7680 × 4320 displays at 60Hz.

AI and Professional Workloads

The RTX PRO 6000 is designed for advanced AI development, generative AI, scientific computing, engineering simulation, real-time visualization, ray-traced rendering, digital content creation and other demanding professional workloads. Its Blackwell architecture combines large-capacity GDDR7 ECC memory with fifth-generation Tensor Cores and fourth-generation RT Cores.

Multi-Instance GPU

The GPU supports NVIDIA Multi-Instance GPU configurations of up to four 24GB instances, two 48GB instances, or one 96GB instance, providing flexible GPU partitioning for supported professional and AI workloads.

Video Engines

PNY specifies four NVENC encoders, four NVDEC decoders and four JPEG engines. The RTX PRO 6000 Blackwell platform also supports modern professional video workflows including advanced AV1 capabilities.

Form Factor and Power

The card measures 5.4 inches in height and 12 inches in length and uses an extra-high-profile, full-length dual-slot design. Maximum power consumption is 600W and the card uses one PCIe CEM5 16-pin power connector. PNY specifies a double-flow-through thermal solution.

Compatibility

• PCI Express 5.0 x16 preferred expansion slot
• Full-height, extra-high-profile dual-slot workstation chassis
• Intel Core i5, i7, i9 or Xeon processor or later
• AMD Ryzen or Epyc-class processor or later
• Windows 10/11 64-bit
• Red Hat Enterprise Linux 7.x
• SUSE Linux Enterprise Desktop 15.x
• Fedora 31
• Ubuntu 18.04
• FreeBSD 11.x
• Solaris 11
• System memory equal to or greater than GPU memory; twice GPU memory recommended
• DisplayPort 2.1 preferred
• HDMI/DVI-compatible displays supported with appropriate adapters
• Internet connection required for driver installation

Package Includes

• PNY NVIDIA RTX PRO 6000 Blackwell Workstation Edition Professional Graphics Card
• One auxiliary power cable (8-pin to dual 6-pin adapter)

Warranty

3-Year Limited Warranty