ASUS ESC NM1-E1 8NVMe
- Condition
- New
- Manufacturer
- ASUS
- Form factor
- 2U
- RAM
- 480GB (LPDDR5x)
- GPU
- 1 x NVIDIA GH200 (Hopper)
- Accelerator model
- NVIDIA GH200 Grace Hopper Superchip (H100 GPU)
- Installed GPU count
- 1
- Maximum GPU count
- 1 integrated GH200 GPU
- Memory per GPU
- 96GB HBM3
- Total GPU memory
- 96GB
- GPU memory bandwidth
- Up to 4000 GB/s (4.0 TB/s)
- GPU form factor
- Grace Hopper Superchip
- GPU interconnect
- NVIDIA NVLink-C2C coherent CPU-GPU interconnect
- GPU-to-CPU interface
- NVLink-C2C at 900 GB/s
- Supported compute data types
- FP64, FP32, TF32, BF16, FP16, FP8, INT8
- GPU virtualization
- GPU passthrough and containerized GPU sharing; support depends on the validated NVIDIA MGX software stack
- MIG support
- Up to 7 isolated GPU instances on the integrated H100 GPU
- Software ecosystem
- NVIDIA CUDA, NVIDIA AI Enterprise and the NVIDIA Grace Hopper software stack
- Cooling
- Air cooling with an ASUS-optimized 2U airflow path for the GH200 module
- Recommended workloads
- Memory-intensive AI and HPC, LLM training and inference, graph neural networks, scientific computing and large-scale analytics
- Dimensions
- 87.5 x 439.5 x 900 mm (H x W x D)
- Weight
- 27 kg
Need a technical consultation on choosing a server?
Submit a requestASUS ESC NM1-E1 8NVMe is a compact 2U enterprise GPU server storage configuration with the eight-NVMe layout listed for this SKU. It is designed for enterprise and research teams that need a space-efficient platform for AI training, inference, HPC, and data analytics.
Key features:
-
Storage basket and service access: The eight-NVMe storage layout provides high-throughput local storage for data-intensive workloads. Service-oriented drive access helps reduce planned maintenance time.
-
Installed accelerator subsystem: The installed accelerator subsystem contains 1 NVIDIA GH200 Grace Hopper Superchip (H100 GPU) accelerators with 96GB of aggregate GPU memory. The configuration is sized for sustained enterprise workloads rather than a general-purpose display role.
-
High-bandwidth data path: The GPU fabric uses NVIDIA NVLink-C2C coherent CPU-GPU interconnect. Storage, GPU and network paths are arranged to support high-throughput data movement and minimize avoidable I/O bottlenecks.
-
Enterprise compute platform: The compute subsystem is configured as follows: CPU: 1 x NVIDIA Grace Hopper Superchip with 72-core Grace CPU; memory: 480GB LPDDR5X CPU memory + 96GB HBM3 GPU memory. This provides CPU, memory and PCIe resources for feeding the accelerator complex and hosting production services.
-
Management and availability: ASUS server management, ASUS Control Center and 1+1 x 2000W 80 PLUS Titanium redundant power supplies support remote administration, planned component replacement and resilient operation in the data center.
Use cases:
-
AI model training and fine-tuning: Uses the installed NVIDIA GH200 Grace Hopper Superchip (H100 GPU) subsystem for deep learning, generative AI and large-model development.
-
Inference and production AI services: Provides GPU memory and parallel compute for LLM, multimodal and enterprise inference workloads.
-
HPC, simulation and analytics: Accelerates scientific computing, numerical simulation, data science and other parallel workloads supported by the NVIDIA GH200 Grace Hopper Superchip (H100 GPU).
-
Data-intensive pipelines: Combines 8NVMe storage configuration listed for this SKU with high-bandwidth GPU and network paths for datasets, checkpoints, staging and local scratch storage.
ASUS ESC NM1-E1 8NVMe is a purpose-built 2U GPU platform for organizations that need 8NVMe storage configuration listed for this SKU, 96GB of installed GPU memory and enterprise-class management in one deployable system.
to Buy a Server
from Servermall
warranty
testing
We want to make sure that our equipment is fully reliable for you. That’s why all our equipment (even new) is subject to complete diagnostics, testing, and part replacement.
testing
We fully load our equipment with various utilities and software. Rest assured, it’ll work for a very long time.
Power supply
The components work at full capacity and are tested for several days. If we are not sure about something, we replace it.
of all equipment
Our regulations imply the same testing for both used and new equipment. Oh, and by the way, we complement the official manufacturer’s warranty with our own warranty.