High Performance AI Inference: Up to 30.3 TFLOPS FP32 and 242 TOPS INT8 performance for AI tasks.
24GB GDDR6 Memory with ECC: Features 24 GB GDDR6 memory with Error Correction Code for data integrity.
PCIe 4.0 x16 Interface: Supports high-speed data transfer via PCIe 4.0 x16 interface.
Low Power Consumption: Power consumption is only 72W with passive cooling solution.
Compact Single-Slot Design: Single-slot, low-profile form factor suitable for space-constrained servers.
Advanced GPU Architecture: Built on NVIDIA Ada Lovelace architecture with 4th Gen Tensor Cores.
Comprehensive Software Support: Compatible with NVIDIA AI Enterprise, CUDA, cuDNN, and TensorRT.
Multi-Use Case Support: Optimized for AI inference, video transcoding, VDI, and lightweight training.