Supermicro Gaudi®2 AI Training Server (SYS-820GH-TNR2)
Building on the success of the original Supermicro Gaudi AI training system, the Gaudi2 AI Training Server prioritizes two key real-world considerations: integrating multi AI training system to analyze diverse AI models faster, while simultaneously multiple scalability function and price advantages.
- GPU: 8 Gaudi2 HL-225H mezzanine cards
- CPU: Dual 3rd Gen Intel® Xeon® Scalable processors
- Memory: 32 DIMMs - up to 8TB registered ECC DDR4-3200MHz SDRAM
- Drives: up to 24 hot-swappable hard drives (SATA/NVMe/SAS)
- Power: 6x 3000W High efficiency (54V+12V)
- Networking: 24x 100GbE (48 x 56Gb) PAM4 SerDes Links by 6 QSFP-DDs
- Expansion Slots: 2x PCI Gen4 switches
- Workloads: AI Training and Inferencing
GAUDI2 PROCESSOR FOR DEEP LEARNING TRAINING AND INFERENCE WORKLOADS.
The Gaudi2 processor significantly increases training and inference performance, building on the high-efficiency, first-generation Gaudi architecture that delivers up to 40% better price performance in the AWS cloud with Amazon EC2 DL1 instances and on-premises with the Supermicro Gaudi®2 AI Training Server.
Gaudi2 delivers training and inference performance on a whole new level with a leap from the first-gen Gaudi 16nm process technology to Gaudi2’s 7nm. It increases the number of AI-customized Tensor Processor Cores–from 8 to 24, adds support for FP8 and integrates a media processing engine for processing compressed media for offloading the host subsystem. Gaudi2’s in-package memory has tripled to 96 GB of HBM2e at 2.45 Terabytes-per-second bandwidth