PNY NVIDIA RTX 4000 Ada Generation 20GB GDDR6 ECC (OEM)
Installment calculator
4,9%
interest
from 963,- monthly
PNY NVIDIA RTX 4000 Ada Generation 20GB GDDR6 ECC (OEM)
PNY NVIDIA RTX 4000 Ada Generation 20GB GDDR6 ECC (OEM) graphics card
The NVIDIA RTX 4000 Ada 20GB is exceptional, offering the performance of a high-end workstation card in a single-slot design with a TDP of just 130W. The Ada Lovelace architecture, with 6144 CUDA cores, 4th generation Tensor Cores supporting FP8, and 3rd generation RT Cores, delivers 26.7 TFLOPS of FP32 performance and up to 327.6 TFLOPS in tensor operations—all without sacrificing an adjacent slot in your setup. With 20GB of GDDR6 ECC memory and a throughput of 360GB/s, even the most demanding scenes, datasets, or AI models won't faze your workstation.
GDDR6 / 20GB
Operating memory
PCIe 4.0 x16
Bus interface
1× 16-pin
Additional power
NVIDIA AD104-400
Chip architecture
650W
Recommended PSU power
Key features of the PNY NVIDIA RTX 4000 Ada Generation 20GB GDDR6 ECC (OEM) graphics card
- Features 20GB of GDDR6 memory with ECC support and 360GB/s throughput on a 160-bit bus, providing ample capacity and reliability for complex CAD models, large rendering scenes, and AI inference tasks
- The NVIDIA Ada Lovelace architecture with 6144 CUDA cores and a boost frequency of up to 1560MHz delivers 26.7 TFLOPS of FP32 performance, with 50% higher performance per core compared to the previous generation
- 4th generation Tensor Cores with FP8 precision support achieve up to 327.6 TFLOPS and deliver up to 3× the AI performance of the Ampere generation—ideal for generative AI, machine learning, and AI inference directly on the workstation
- 3rd generation RT Cores with 61.8 TFLOPS of performance enable real-time ray tracing for photorealistic visualisations, product design, and architectural presentations directly in the viewport
- The single-slot design with a 130W TDP is exceptional in the 20GB professional card segment, fitting into builds where dual-slot cards face space constraints
- Four DisplayPort 1.4a outputs supporting resolutions up to 7680×4320px let you connect four monitors simultaneously for a highly flexible multi-window work environment
- Two 8th generation hardware NVENC encoders with AV1 format support ensure fast, high-quality video encoding in 4K and 8K, with AV1 encoding being 40% more efficient than H.264
- The PCI Express 4.0 x16 interface, with double the throughput of PCIe Gen 3, eliminates data bottlenecks when transferring large datasets and 3D scenes from CPU memory
20GB of ECC memory for error-free work without compromise
In a professional setting, it’s not enough for a card to be fast—it must also be reliable. That’s why the NVIDIA RTX 4000 Ada 20GB comes with GDDR6 memory with ECC support, which actively detects and corrects errors that occur during computations. For architects working with complex BIM models, engineers running simulations, or researchers training AI models, this means one thing: results you can trust. With 20GB of memory and a throughput of 360GB/s, even the most demanding scenes, datasets, and multi-application workflows won't hit memory limits—and the single-slot design with a 130W TDP handles it all without taking over your entire system.
Ada Lovelace for creatives and developers
Modern workflows aren't limited to a single discipline—and the NVIDIA RTX 4000 Ada 20GB is designed with that in mind. 4th generation Tensor Cores with FP8 support deliver up to 3× the AI performance of the previous generation, paving the way for generative AI, Stable Diffusion acceleration, and AI tools directly within creative applications. 3rd generation RT Cores with 61.8 TFLOPS of performance provide real-time photorealistic ray tracing for product design, architectural visualisations, and VFX scenes. Two hardware AV1 encoders take the load off the CPU when exporting 4K and 8K video—and with DLSS 3 technology, you get extra smoothness in situations where it was previously only possible with more expensive hardware.
Specifications
Dimensions
AI
Chip
Performance & usage
Memory
Frequency
Interface
Technology
Cooling system
Consumption
Tops
AI specifications
PNY Technologies Europe, ZAC du Phare 9 rue Joseph Cugnot 33708 Mérignac cedex, France https://www.pny.com/en-eu/support-footer/technical-support
This product is not an independently functional unit and may require professional installation.
Specifications can be changed without notice. Images are for illustrative purposes only.
