What memory capacity and type does the NVIDIA 900-5G153-2500-000 provide?

It features 96GB GDDR7 memory with ECC on a 512-bit bus, delivering 1792 GB/s bandwidth. This allows loading massive 3D scenes, AI models, or simulation datasets entirely in GPU memory, eliminating PCIe transfer latency for smoother workflows in professional applications.

How many CUDA, Tensor, and RT cores are in this GPU?

The RTX PRO 6000 Blackwell Max-Q has 24,064 CUDA cores, 752 5th-generation Tensor Cores, and 188 4th-generation RT Cores. These enable 110 TFLOPS single-precision, 333 TFLOPS RT performance, and 3511 AI TOPS for AI training, ray-traced rendering, and compute tasks.

What is the power consumption and form factor?

Total board power is 300W via 1x 16-pin PCIe CEM5 connector, with dual-slot 4.4" H x 10.5" L full-height form factor and active cooling. Max-Q design balances elite performance with efficiency for compact workstations without thermal throttling.

Which displays and resolutions are supported?

Includes 4x DisplayPort 2.1 outputs supporting up to 7680x4320@60Hz per display and 4 simultaneous 8K monitors. Supports HDR, AV1 encode/decode for video production and multi-app workflows.

What APIs and compute platforms are compatible?

Supports DirectX 12 Ultimate, OpenGL 4.6, Vulkan 1.3, CUDA 12.8, OpenCL 3.0, DirectCompute, Shader Model 6.6. Optimized for professional software like Autodesk, Adobe, Siemens NX, Ansys.

Is this GPU suitable for mobile or compact workstations?

Yes, Max-Q edition optimizes for thinner, quieter systems with 300W TDP while retaining full Blackwell capabilities. Compatible with Intel Xeon, AMD EPYC/Ryzen workstations via PCIe 5.0 x16.

900-5G153-2500-000 - NVIDIA 900-5G153-2500-000 96GB GDDR7 24K CUDA PCIe5 GPU

Name: NVIDIA 900-5G153-2500-000 96GB GDDR7 24K CUDA PCIe5 GPU
Brand: NVIDIA
SKU: 900-5G153-2500-000

900-5G153-2500-000 - NVIDIA 900-5G153-2500-000 96GB GDDR7 24K CUDA PCIe5 GPU | US IT Hardware

Product overview

The NVIDIA RTX PRO 6000 Blackwell Max-Q Workstation Edition (900-5G153-2500-000) is a dual-slot professional GPU built on Blackwell architecture, delivering extreme performance for AI-driven workflows, advanced 3D rendering, simulations, and large-scale data processing. It features 24,064 CUDA cores, 752 5th-generation Tensor Cores for FP4/DLSS 4 AI acceleration, 188 4th-generation RT Cores, and 96GB GDDR7 ECC memory on a 512-bit bus providing 1792 GB/s bandwidth. With PCIe 5.0 x16 interface, 300W TDP via 16-pin connector, active dual-flow cooling, and 4x DisplayPort 2.1 outputs supporting 8K@60Hz, it enables massive datasets to stay in GPU memory, eliminating CPU swaps for seamless operation. Optimized for professional workstations, it supports DirectX 12 Ultimate, Vulkan 1.3, CUDA 12.8, OpenCL 3.0. Ideal for CAD, DCC, scientific visualization, VR/AR, neural rendering, and local AI model training in compact, power-efficient form factors.

96GB GDDR7 ECC memory with 1792 GB/s bandwidth handles massive datasets in GPU.
24,064 CUDA cores and 752 5th-gen Tensor Cores deliver 3511 AI TOPS performance.
PCIe 5.0 x16 interface and 300W TDP enable high-bandwidth workstation integration.
4x DisplayPort 2.1 outputs support 8K@60Hz multi-display professional workflows.
512-bit memory bus and active cooling ensure sustained peak performance.
ECC protection and Blackwell architecture provide reliability for mission-critical tasks.

Expedited and overnight shipping options are available at checkout, priced by destination and service.
Purchase orders accepted from qualified institutions.
Refunds within 14 days and replacements within 30 days of delivery with an RMA (25% restocking fee on customer-initiated returns), and every part carries our Standard 30-Day Warranty - see our return policy or contact US IT Hardware.

Featuring

Manufacturer part #	900-5G153-2500-000
Brand	NVIDIA
Condition	New
Model	RTX PRO 6000 Max‑Q
Cooling	Active double-flow-through
Product Line	RTX PRO Blackwell
Product Type	Graphic Card
Supported APIs	DirectX 12, OpenGL 4.6, Vulkan 1.3, CUDA 12.8, OpenCL 3.0

Storage

Memory Bus
Memory	96 GB GDDR7 ECC

Compatible systems

Optimized for PCIe 5.0 workstations with Intel Core/Xeon or AMD Ryzen/EPYC CPUs, 850W+ PSUs. Full NVIDIA driver support on Windows/Linux.

Dell Precision
Model: 7000 Series
Form factor: Workstation
Generation: Latest
PCIe 5.0 x16 slot required for full bandwidth.
HP Z
Model: Z8/Z4 G5
Form factor: Desktop
Generation: Current
300W+ PSU with 16-pin adapter.
Lenovo ThinkStation
Model: P Series
Form factor: Tower
Generation: Gen 3+
Dual-slot clearance needed.

Best for

AI/ML Workloads
Local fine-tuning of large AI models with 96GB memory and 5th-gen Tensor Cores supporting FP4 precision.
3D Rendering & Visualization
Real-time ray tracing and neural rendering for complex scenes using 4th-gen RT Cores and RTX Neural Shaders.
Scientific Simulations
High-fidelity simulations in CFD, molecular dynamics with 24,064 CUDA cores and ECC for data integrity.

Details

Shipping, warranty & returns

Customer reviews

No reviews yet — be the first to review this part.

Dimensions	4.4in H x 10.5in L
Form Factor	4.4” (H) x 10.5” (L) dual slot

NVIDIA 900-5G153-2500-000 96GB GDDR7 24K CUDA PCIe5 GPU

Product overview

Featuring

Storage

Compatible systems

Best for

Details

Shipping, warranty & returns

Customer reviews

Processor

Power Consumption

Physical Characteristics

Display

Interface/Ports

Frequently asked questions

AI TOPS	3511
RT Cores	188 (4th Gen)
CUDA Cores	24,064
Tensor Cores	752 (5th Gen)
RT Performance	333 TFLOPS
GPU Architecture	NVIDIA Blackwell
Single-Precision	110 TFLOPS