- GPU Memory: 48GB GDDR6 ECC for AI, HPC, and data-intensive workloads
- CUDA Cores: 10,752 CUDA cores for high parallel compute performance
- RT Cores: 84 2nd-Gen RT Cores for ray tracing and rendering workloads
- Tensor Cores: 336 3rd-Gen Tensor Cores for accelerated AI training and inference
- Performance: 37.4 TFLOPS FP32 | 149.7 TFLOPS FP16/BF16 Tensor | 299.4 TFLOPS with sparsity acceleration
- Interface: PCIe 4.0 x16 for high-bandwidth data center connectivity
- Display Ports: 3x DisplayPort 1.4 for multi-display, high-resolution workstation output
- Passive Cooling: Fanless design for efficient airflow cooling in data center and multi-GPU server environments
- Form Factor: Dual-slot, full-height, full-length passive-cooled design
- Power Consumption: 300W TDP | 8-pin power connector
- Enterprise Features: Supports CUDA, cuDNN, TensorRT, and NVIDIA AI software stack | vGPU ready — NVIDIA RTX Virtual Workstation and Virtual Compute Server supported