NVIDIA RTX PRO 4500 Blackwell Server Edition Passive 32GB

137,999,-

NVIDIA RTX PRO 4500 Blackwell Server Edition Passive 32GB

ℹ️

This product is intended solely for businesses and educational institutions. Completing an order on the e-shop does not constitute a binding contract, but rather the submission of an enquiry.

The sale of this product is subject to export control, sanctions and other legal restrictions. Before an order can be accepted, the buyer, the end user, the country of delivery and the intended use must be verified. An order can only be accepted after successful authorisation by the supplier.

If the required conditions are not met, the enquiry may be declined or the order cancelled without any entitlement to delivery of the goods.

If you are interested in purchasing an AI solution or accelerator, you can also email us at [email protected]. We will then prepare a proposal and an individual contract for you. For further information, please refer to our terms and conditions (Articles 3.33 and 3.34).

NVIDIA logo

NVIDIA RTX PRO 4500 Blackwell Server Edition Passive 32GB graphics card

The NVIDIA RTX PRO 4500 32GB Blackwell delivers 5th generation Tensor Cores with FP4 support, 4th generation RT Cores with RTX Mega Geometry, and 32GB of GDDR7 ECC memory with 800GB/s bandwidth in a dual-slot, full-height design with a TBP of just 200W. With 1617 AI TOPS, 51 TFLOPS in FP32, and native support for NVIDIA NemoClaw for agentic AI with security rules, the NVIDIA RTX PRO 4500 32GB Blackwell goes a step beyond the previous Ada Lovelace generation – and with significantly lower power requirements than higher-end Blackwell series models.

NVIDIA GeForce RTX 40

GDDR7 / 32GB

Operating memory
NVIDIA GPU

PCIe 5.0 x16

Bus interface
NVIDIA GPU

1× PCIe CEM5 16-pin

Auxiliary power
NVIDIA GeForce RTX 40

Blackwell (GB203)

Chip architecture
NVIDIA GeForce RTX 40

650W

Recommended PSU wattage
NVIDIA RTX PRO 4500 Blackwell Server Edition Passive 32GB

Key features of the NVIDIA RTX PRO 4500 Blackwell Server Edition Passive 32GB graphics card

  • 32GB of GDDR7 ECC memory with 800GB/s bandwidth on a 256-bit bus for smooth work with large AI models and complex 3D scenes without memory limits
  • NVIDIA Blackwell architecture with 10496 CUDA Cores and 5th generation Tensor Cores achieves 1617 AI TOPS and 51 TFLOPS in FP32, while FP4 precision enables faster processing of AI models with lower memory requirements for local LLM fine-tuning and generative AI
  • 4th generation RT Cores with 153 TFLOPS performance and RTX Mega Geometry technology process up to 100× more ray-traced triangles than the previous generation for physically accurate real-time visualisations and engineering simulations
  • New Streaming Multiprocessors with Neural Shaders integrate neural networks directly into programmable shaders for next-generation AI-augmented graphics workflows
  • NVIDIA NemoClaw support enables the secure deployment of autonomous agentic AI systems with strict privacy guarantees and security guardrails for local AI workflows with sensitive data
  • Two 9th generation NVENC video encoders and two 6th generation NVDEC decoders with support for 4:2:2 10-bit H.264 and HEVC enable real-time professional video processing for production and VFX workflows
  • Powered by a 1× PCIe CEM5 16-pin connector with a 165W TBP, making it one of the most efficient professional Blackwell cards – the recommended PSU for a workstation build is at least 650W
  • Equipped with passive cooling in a single-slot FHFL design with a 165W TBP, which fully integrates into the server chassis airflow and ensures quiet, maintenance-free operation for production data centres, edge and cloud deployments
  • Features fanless passive cooling that fully integrates into the server chassis airflow, ensuring quiet, maintenance-free operation and eliminating fans as a potential point of failure in a production data centre

1617 AI TOPS for local agentic AI with data protection

Agentic AI systems – autonomous agents that plan, decide, and act – place new demands on hardware: not just performance, but also data security and privacy. The NVIDIA RTX PRO 4500 32GB Blackwell natively supports NVIDIA NemoClaw, an open-source reference stack that adds security and privacy guardrails to local AI agents. With 1617 AI TOPS and 5th generation Tensor Cores with FP4 support, the card handles both inference and fine-tuning of local LLMs without cloud dependency. And your sensitive data remains on your workstation, under your control, at all times.

NVIDIA GPU

51 TFLOPS and RTX Mega Geometry: next-generation visualisation even in the mid-range

Neural Shaders and RTX Mega Geometry are no longer exclusive to flagship GPUs. The NVIDIA RTX PRO 4500 32GB Blackwell brings 4th generation RT Cores with 153 TFLOPS performance and the ability to process up to 100× more ray-traced triangles than the previous generation directly to the more accessible professional segment. For architects working with detailed BIM models, manufacturing designers validating prototypes in a virtual environment, or VFX artists rendering photorealistic scenes, this means a computational leap that shortens iteration cycles and expands the boundaries of visualisation without needing to reach for the flagship model.

NVIDIA GPU

Blackwell generation performance without demanding power requirements

Next-generation professional hardware doesn't have to mean overhauling your entire workstation infrastructure. With a 165W TBP, the NVIDIA RTX PRO 4500 32GB Blackwell is one of the most efficient professional Blackwell cards – yet it delivers 32GB of GDDR7 ECC memory with 896GB/s bandwidth, DLSS 4 Multi Frame Generation, and ISV certifications for key professional applications. For organisations building AI-ready workstation fleets or IT managers planning an upgrade without needing to replace power supplies, the RTX PRO 4500 Blackwell is the natural choice – Blackwell generation performance at a reasonable power cost.

NVIDIA GPU
NVIDIA GeForce RTX 40

NVIDIA Blackwell architecture

A breakthrough generation for AI, neural rendering, and scientific computing
Nvidia GeForce RTX 40

5th generation Tensor Cores

1617 AI TOPS with FP4 for local LLM and generative AI workflows
Nvidia GeForce RTX 40

32GB GDDR7 ECC

800GB/s for large AI models and complex 3D scenes
Nvidia GeForce RTX 40

New streaming multiprocessors

Neural shaders for next-generation AI-augmented graphics workflows
Nvidia GeForce RTX 40

4th generation RT cores

RTX Mega Geometry for physically accurate scenes with billions of polygons
Nvidia GeForce RTX 40

NVIDIA NemoClaw

Secure deployment of agentic AI with privacy guardrails for sensitive data
Nvidia GeForce RTX 40

AI Management Processor

Intelligent distribution of computing resources for professional workloads
Nvidia GeForce RTX 40

PCIe 5.0 x16

Double the throughput of PCIe Gen 4 for data-intensive AI tasks

Specifications

Chip

Chip manufacturer NVIDIA
Model designation RTX PRO 4500
Graphics processor GB203
Chip architecture Blackwell
Number of stream processors 10,496×
Production technology 4 nm

Performance & usage

Use Professional
Card profile FHFL

Memory

VRAM capacity 32 GB
Memory type GDDR7
Bus width 256-Bit

Frequency

Core frequency 1,590 MHz
Core clock speed (Boost clock) 2,617 MHz
Memory frequency 28,000 MHz (28 GHz)
Effective memory clock (bandwidth) 28 Gbps

Interface

PC connection PCI Express x16 5.0
Bus interface PCIe 5.0 x16

Technology

DirectX version 12 Ultimate

Cooling system

Cooling type Passive
Cooling size One-slotted

Dimensions

Width 266.7 mm
Depth 111.8 mm

Consumption

TDP 165 W

AI

Total number of TOPS 1,617
Supported precision FP32

NCNR (Non-Cancellable, Non-Returnable)

Způsob nákupu pro firmy Na poptávku s individuální smlouvou

Tops

Total number of TOPS 1,617

AI specifications

Primary AI use LLM Inference
Code:  NVrpbs03
Warranty: 36 months
Links: Producer's Website: