900-2G153-0100-000 - NVIDIA 900-2G153-0100-000 96GB GDDR7 passive server GPU

Name: NVIDIA 900-2G153-0100-000 96GB GDDR7 passive server GPU
Brand: NVIDIA
SKU: 900-2G153-0100-000

900-2G153-0100-000 - NVIDIA 900-2G153-0100-000 96GB GDDR7 passive server GPU | US IT Hardware

Product overview

The NVIDIA RTX PRO 6000 Blackwell Server Edition is a datacenter GPU designed as a PCI Express 5.0 x16 graphics solution for server deployments. It is equipped with 96 GB of GDDR7 memory with ECC on a 512 bit interface, delivering 1597 GB per second of memory bandwidth to handle data intensive workloads. The GPU integrates 24,064 CUDA parallel processing cores, 752 fifth generation NVIDIA Tensor Cores, and 188 fourth generation NVIDIA RT Cores to support compute, AI, and ray tracing tasks. Single precision FP32 performance is specified at 120 TFLOPS, with peak FP4 AI performance of 4 PFLOPS and RT core performance of 355 TFLOPS. The card uses a passive thermal solution in a dual slot form factor measuring 4.4 inches high by 10.5 inches long and supports up to 600 watts configurable power via a PCIe CEM5 16 pin power connector.

96 GB GDDR7 ECC with 1597 GB per second bandwidth
24,064 CUDA cores plus 752 fifth gen Tensor Cores
120 TFLOPS FP32 and 4 PFLOPS peak FP4 AI performance
188 fourth gen RT Cores with 355 TFLOPS RT performance
PCI Express 5.0 x16, dual slot, passive thermal design
Configurable power consumption up to 600 watts

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

Product Type
Manufacturer part #	900-2G153-0100-000
Brand	NVIDIA
Condition	New
Model	RTX PRO 6000 Blackwell Server Edition
Edition	Server Edition
CUDA Cores	24064
Board Power	600 W
Form Factor	Dual slot
Memory Type	GDDR7 ECC
Cooling Type	Passive
Product Line	NVIDIA RTX PRO

Compatible systems

Designed as a passive RTX Pro 6000 Blackwell GPU for Dell PowerEdge and other server platforms.

Dell PowerEdge
Form factor: Server
Listed as DELL Nvidia Rtx Pro 6000 Blackwell Server Edition Passive GPU 96GB 900-2G153-0100-000-DELL on IT Creations.

Best for

Datacenter GPU
Suited for datacenter server installations with a PCI Express 5.0 x16 interface, dual slot form factor, and passive thermal solution supporting up to 600W configurable power.
AI and compute
Useful for AI and compute workloads that benefit from 24,064 CUDA cores, 752 fifth generation Tensor Cores, and 120 TFLOPS of single precision FP32 performance.
Visual computing and ray tracing
Applies to visual workloads that can leverage 188 fourth generation RT Cores and 355 TFLOPS RT core performance in combination with 96 GB of GDDR7 ECC memory.
Memory intensive applications
Appropriate for memory intensive applications that require 96 GB of GDDR7 with ECC on a 512 bit interface and 1597 GB per second of memory bandwidth.
Multi instance GPU environments
Fits environments that need up to 4 multi instance GPU partitions at 24 GB each for workload isolation and resource sharing on a single physical GPU.

Details

Shipping, warranty & returns

Customer reviews

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

Card Dimensions	4.4 in height, 10.5 in length

GPU Architecture	NVIDIA Blackwell

GPU Memory Capacity	96 GB

Maximum Power Consumption	600 W

Bus Interface	PCI Express 5.0 x16
Connector Type	PCI Express
Power Connector	1x PCIe CEM5 16-pin
Memory Interface Width	512 bit

Frequently asked questions

What is the memory configuration of the NVIDIA RTX PRO 6000 Blackwell Server Edition?

The GPU is equipped with 96 GB of GDDR7 memory with ECC support. It uses a 512 bit memory interface and provides 1597 GB per second of memory bandwidth. This combination allows the card to handle large datasets and memory intensive workloads while providing error correcting capabilities for improved reliability.

Which interface and form factor does this GPU use for server integration?

The NVIDIA RTX PRO 6000 Blackwell Server Edition connects via a PCI Express 5.0 x16 graphics bus. It is a dual slot card with physical dimensions of 4.4 inches in height by 10.5 inches in length and uses a passive thermal solution. These characteristics are important for confirming compatibility with server chassis and airflow designs.

How much power does the RTX PRO 6000 Blackwell Server GPU require?

Power consumption for this GPU is specified as up to 600 watts and is configurable. It draws power through a single PCIe CEM5 16 pin power connector. When planning deployment, datacenter teams should account for the maximum power rating in rack power budgets and ensure compatible power delivery infrastructure.

What compute and AI performance metrics does this GPU provide?

The card delivers 120 TFLOPS of single precision FP32 performance and 4 PFLOPS of peak FP4 AI performance. It includes 24,064 CUDA parallel processing cores and 752 fifth generation NVIDIA Tensor Cores to support compute and AI workloads. These metrics help evaluate suitability for simulation, inference, and other GPU accelerated tasks.

Does the RTX PRO 6000 Blackwell Server Edition support multi instance GPU operation?

Yes, the GPU supports up to four multi instance GPU configurations, each with 24 GB of memory. This allows a single physical GPU to be partitioned into multiple logical instances for workload isolation, resource sharing, and more flexible utilization within a datacenter environment.

What ray tracing capabilities does this server GPU offer?

The GPU is equipped with 188 fourth generation NVIDIA RT Cores and provides 355 TFLOPS of RT core performance. These hardware ray tracing resources are designed to accelerate workloads that make use of ray traced rendering or other RT core accelerated operations alongside traditional compute and AI tasks.

What cooling solution is used on the RTX PRO 6000 Blackwell Server Edition?

This model uses a passive thermal solution in a dual slot design. There are no onboard fans, so it relies on system level airflow within the server chassis for cooling. This approach is typical in datacenter environments and requires appropriate airflow design to maintain reliable operating temperatures.

NVIDIA 900-2G153-0100-000 96GB GDDR7 passive server GPU