Four NVIDIA RTX Ada and L40s GPUs deliver exceptional parallel processing power for large language model training and inference
192GB combined VRAM enables efficient handling of 70 billion parameter FP16 models and fine-tuning
Compact 2U rackmount design saves space in professional and data centers
RAID support ensures data redundancy and reliability for mission-critical AI tasks
PCIe expansion slots allow for additional GPU or accelerator installation
Summarized by Shop
Designed for AI researchers and data scientists, this compact 2U rackmount server supports up to four NVIDIA GPUs, making it ideal for fine-tuning and inference with large language models. With support for NVIDIA RTX Ada and L40S graphics cards, it delivers the power needed for demanding AI workloads.
Featuring up to 192GB of combined VRA