HPC AI Server APY SCG 5U 8 GPU NVIDIA HGX H200 SXM Intel Xeon Scalable

The APY Compute Server is a 5U server designed for generative AI and high-performance computing (HPC). It is an NVIDIA-certified system that integrates the latest technologies to deliver exceptional performance and scalability.

Processors

This server supports two 5th or 4th Generation Intel Xeon Scalable processors, offering up to 64 cores and 128 threads per processor. The Intel Xeon CPU Max series is also supported, which allows for a maximum computing power of 350W per processor.

Memory

The server has 32 DIMM slots for DDR5 RDIMM memory, with a maximum capacity of 4TB per processor. This configuration enables smooth running of memory-intensive applications.

GPU

The APY server is equipped with 8 NVIDIA HGX H200 GPUs with SXM connectors. These GPUs deliver breakthrough performance for AI and HPC, with 900 GB/s GPU-to-GPU bandwidth thanks to NVLink and NVSwitch technologies.

Storage

The server has 8 2.5-inch SATA/SAS/NVMe hard drive bays. Support for RAID 0, 1, 5, 10 and JBOD helps ensure data security and availability.

Network

It is equipped with 2 x 10GbE Ethernet ports and 1 Gigabit Ethernet port. This configuration enables high-speed network connectivity to meet the needs of the most demanding applications.

Others

The server is powered by a 3000W redundant power supply and has an optimized air cooling system to ensure reliable and stable operation.

Highlights

High performance for generative AI and HPC Large memory and storage capacity High GPU-to-GPU bandwidth Cooling-optimized design NVIDIA certified

This server is an ideal choice for data centers that need a powerful and versatile server for generative AI and HPC.

Technical specifications Processors: 2x 5th or 4th Gen Intel Xeon Scalable Memory: 32x DDR5 RDIMM DIMMs, up to 4TB per processor GPU: 8x NVIDIA HGX H100 SXM5 Storage: 8x 2.5" SATA/SAS/NVMe bays Network: 2x 10GbE, 1x GbE Power supply: 2x 3000W redundant Cooling: Air