8 H100

Co-located in a lambda colorado datacenter

Operating system: Ubuntu 22.04: Includes Lambda Stack for managing TensorFlow, PyTorch, CUDA, cuDNN, etc.

Processor: 2x AMD EPYC 9654: 96 cores, 2.4~3.7GHz, 384MB cache, PCIe 5.0

GPUs: 8x NVIDIA H100 (80GB) SXM5 - Education/Inception Discount: HGX platform with NVLink and NVSwitch fabric

System memory: 1536 GB: DDR5-4800 ECC RDIMM

Drives: 2x 1.92 TB M.2 NVMe: Data center SSD, 1 DWPD, PCIe 4.0

Standard Networking: 2x 10 Gbps RJ45 Ethernet ports, 1x dedicated IPMI port