We take cryptocurrency — so the coins you hold can buy real hardware.Pay in crypto, spend it on real hardware. Free insured shipping worldwide over $399. Plain, unbranded boxes, sent from Germany.Free insured shipping over $399. Your rate is locked for 30 minutes once checkout opens.Rate locked 30 minutes. Every item sealed, serial-checked and covered by a 24-month warranty.Sealed, 24-month warranty.

Saved items Sign in
1U barebone server chassis with integrated NVIDIA GH200 Grace Hopper Superchip

Supermicro

Supermicro ARS-111GL-NHR barebone server GH200 576GB 96GB HBM3 480GB LPDDR5X

SKU CH-Z79G-XVM5D MPN ARS-111GL-NHR

1U barebone server with Grace Hopper Superchip, 96GB HBM3, 480GB LPDDR5X, 576GB coherent memory, and 900GB/s CPU-GPU interconnect

  • Integrated Grace Hopper Superchip accelerates AI workloads
  • 576GB coherent memory capacity supports large language models
  • 96GB HBM3 GPU memory delivers high-bandwidth compute
  • 480GB LPDDR5X CPU memory enables massive datasets
  • NVLink C2C interconnect provides 900GB/s CPU-GPU bandwidth

GH200 Grace Hopper Superchip

This 1U barebone server integrates the NVIDIA GH200 Grace Hopper Superchip, combining an ARM-based Grace CPU with an H100 GPU. It is built for AI and high-performance computing workloads that demand massive coherent memory and tight CPU-GPU coupling. The platform targets researchers and engineers training large language models.

NVLink-C2C 900GB/s interconnect

The chip-to-chip link delivers 900GB/s bandwidth between the Grace CPU and the H100 GPU, removing the PCIe bottleneck for unified memory access. Two PCIe 5.0 x16 slots accept NVIDIA BlueField-3 or ConnectX-7 adapters for high-speed networking or storage offload. The system supports two E1.S drives connected directly to the processor.

576GB coherent memory

A single node presents up to 576GB of coherent memory, split between 480GB of LPDDR5X system memory and 96GB of HBM3 on the GPU. This capacity decides whether the server can hold the target model weights in one node. Nine hot-swap fans with optimal speed control manage thermal load in the dense 1U chassis.

Highlights

  • Integrated Grace Hopper Superchip accelerates AI workloads
  • 576GB coherent memory capacity supports large language models
  • 96GB HBM3 GPU memory delivers high-bandwidth compute
  • 480GB LPDDR5X CPU memory enables massive datasets
  • NVLink C2C interconnect provides 900GB/s CPU-GPU bandwidth

Specifications

No published specifications for this item yet.

Questions about this item

Will the 9 Hot-Swap Heavy Duty Fans need periodic replacement?

The nine hot-swap heavy-duty fans are designed for tool-less removal, so individual units can be replaced without shutting down the server.

Which component hits its limit first during sustained all-day workloads?

The 900 GB/s NVLink C2C link between the Grace CPU and H100 GPU saturates before the 576 GB coherent memory pool, making interconnect bandwidth the first bottleneck.

How much power does the system draw and how loud is it when running continuously?

With nine heavy-duty fans under optimal speed control, the chassis moves high airflow to cool the integrated Grace Hopper Superchip, keeping noise proportional to thermal load.

New to this? Buying server hardware with crypto — what to check, what it costs to pay, and what is never asked for.

Also in this aisle

People compared these

Same shelf, same checkout — eight coins and a 30-minute rate lock.