The NVIDIA RTX 4000 ADA is a high-end workstation GPU. While using older memory generation GDDR6, it still provides the speed and power for most large language models with its 20GB. Engineered for AI professionals, engineers, and businesses dependent on AI & LLM training, this card has high compute power, efficiency, reliability and built on Ada Lovelace technology that is connected with consumer RTX GPUs. This card with software projects such as LM Studio or Abacus AI, you can run the most efficient and high quantized models, such as some smaller contained models from DeepSeek, GPT among others at the highest token speeds. Key Features: NVIDIA RTX 4000 ADA 20GB Brand NVIDIA Model # RTX 4000 ADA GPU Memory 20GB Memory Generation GDDR6 CUDA Cores 6144 Memory Bandwidth 360GB/s Interface PCIe 4.0 x16 Tensor Cores 192 Ray Tracing Cores 48 Power 130W Active Dimensions Single Slot FH Display Capabilities 4x Mini Display Ports 1.4 MPN 900-5G190-2270-000