The 7u APY server is a high -end solution designed for demanding artificial intelligence applications (IA), high performance calculation (HPC) and massive data analysis. It is distinguished by its computing power, its energy efficiency and its advanced connectivity.
1. Exceptional calculation performance
equipped with two Intel® Xeon® Scalable processors of 5ᵉ generation, the server supports a maximum TDP of 350 W per socket, thus offering a high treatment capacity adapted to intensive workloads in calculation and IA.
2. Advanced GPU acceleration with NVIDIA HGX H200
It incorporates the NVIDIA® HGX H200 system with eight GPU Tensor Core, to achieve unmatched performance for training and inference of artificial intelligence models.
3. Ultra-fast GPU interconnection
thanks to NVLink and a GPU GPU interconnection offering a bandwidth of 900 GB/S, the server provides ultra-fast communication between GPUs, thus improving the treatment of IA and HPC workloads.
4. Designed for generative AI and large models (LLM)
APY optimized this server for generative AI applications and deep learning, with specialized support for large (LLM) language models and dedicated software infrastructure.
5. Advanced energy efficiency and cooling
It has independent air flow tunnels for CPU and GPU, thus optimizing cooling and reducing energy consumption. It is equipped with a redundant 4+2 or 3+3+3 power supply, 80 Plus® Titanium certified, guaranteeing optimal energy efficiency.
6. Exceptional connectivity and extensibility
The server offers:
12+1 PCIe locations,
32 DIMM locations for an ultra-fast DDR5 memory,
10 bays nvme for high speed storage,
Double connectivity LAN 10GB, with extension options for advanced network needs.
7. Reduced latency and GPU-Stockage Optimization
The optimized proximity of network cards (NIC) and storage compared to GPUs, as well as the 1: 1 GPU/NIC ratio, allow minimum latency and improved performance, in particular thanks to Nvidia Gpudirect Storage.
The APY server is an ideal platform for companies and data centers wishing to deploy IA and HPC solutions. Its powerful architecture, its energy efficiency and its advanced interconnection make it an essential choice for the most demanding workloads.