PNY NVIDIA RTX 6000 Ada Generation 48GB GDDR6 ECC
PNY NVIDIA RTX 6000 Ada Generation 48GB GDDR6 ECC
PNY NVIDIA RTX 6000 Ada Generation 48GB GDDR6 ECC
The PNY NVIDIA RTX 6000 Ada Generation graphics card is designed for professional workstations, handling tasks in artificial intelligence and demanding 3D graphics. This graphics card is built on the Ada Lovelace architecture and integrates 48GB of GDDR6 ECC memory with a bandwidth of up to 960GB/s. It utilises 18176 CUDA cores and an active dual-slot blower-style cooler for heat dissipation. The card connects to the motherboard via a PCI Express 4.0 ×16 interface and requires one 16-pin PCIe CEM5 connector for additional power.
GDDR6 / 48GB
VRAM
PCIe Express 4.0 ×16
Bus interface
16-pin
Additional power supply
Ada Lovelace
Chip architecture
Key features of the PNY NVIDIA RTX 6000 Ada Generation 48GB GDDR6 ECC graphics card
- Ada Lovelace architecture core with 18176 CUDA units
- 48GB of GDDR6 memory with ECC support
- 142 third-generation RT cores and 568 fourth-generation Tensor cores
- Power consumption of 300W under full load
- PCIe 4.0 ×16 interface for fast data transfer
- Four DisplayPort 1.4a connectors for video output
Computing power of the Ada Lovelace architecture
The AD102 GPU, manufactured using a 4nm process, contains 76.3 billion transistors. The architecture delivers double the speed for single-precision FP32 floating-point operations. Users in 3D CAD and CAE benefit from high throughput for complex simulations and modelling. The card demonstrates stable operation even under sustained 100% load.
PNY NVIDIA RTX 6000 Ada Generation graphics card and rendering
The third-generation RT cores accelerate the creation of photorealistic visualisations and virtual prototypes. The card includes 142 RT cores with double the throughput compared to the previous generation. These elements speed up real-time ray tracing calculations without compromising smoothness. This allows professionals to work with accurate representations of materials and shadows.
Artificial intelligence and Tensor cores
The fourth-generation Tensor cores introduce support for the FP8 data format for machine learning. A total of 568 Tensor cores provide more than double the AI performance for inference tasks. The card processes neural networks with high efficiency and reduces the time needed to train models. The system is therefore ideal for data science and advanced research. The card features AV1 encoding via the eighth-generation NVENC encoder. This codec shows 40% greater efficiency compared to the older H.264 format. The hardware supports virtualisation with NVIDIA RTX Virtual Workstation for performance sharing. Four DisplayPort 1.4a connectors provide connectivity to monitors.
Technical specification
- Model name: RTX 6000
- Chip architecture: Ada Lovelace
- VRAM capacity: 48GB GDDR6 ECC
- Bus width: 384-bit
- Number of CUDA cores: 18176
- PC interface: PCI Express ×16 4.0
- Display connectors: 4× DisplayPort 1.4a
- Additional power: 1× PCIe CEM5 16-pin
- Power consumption: 300W
Specifications
Chip
Performance & usage
Memory
Frequency
Interface
Technology
Cooling system
Consumption
AI
Tops
PNY Technologies Europe, ZAC du Phare 9 rue Joseph Cugnot 33708 Mérignac cedex, France https://www.pny.com/en-eu/support-footer/technical-support
Specifications can be changed without notice. Images are for illustrative purposes only.
