ecs.ebmgn9t.48xlarge by Alibaba Cloud
(All-cores)
(Single-core)
Specifications
Server Metadata
Vendor ID | alicloud |
Name | ecs.ebmgn9t.48xlarge |
Description | ecs.ebmgn9t family (192 vCPUs, 1536 GiB RAM, 0 GB storage, 8xG59 32768 GiB VRAM) |
Family | ecs.ebmgn9t |
Hw Virt | - |
Status | inactive |
Observed At | 2026-07-25T19:47:22.775375 |
Availability
| REGION / ID | SPOT | ONDEMAND |
|---|
Processor
vCPUs | 192 |
Hypervisor | KVM |
CPU Allocation | Dedicated |
CPU Cores | 192 |
CPU Speed | 2.6 GHz |
CPU Architecture | x86_64 |
CPU Manufacturer | Intel |
CPU Family | Xeon |
CPU Model | 8575C |
System Resources and Accelerators
| MEMORY | |
|---|---|
Memory Amount | 2 TB |
| GPU | |
|---|---|
GPU Count | 8 |
GPU Memory Min | 32 GiB |
GPU Memory Total | 256 GiB |
GPU Model | G59 |
GPUs |
| STORAGE | |
|---|---|
Storage Size | 0 GB |
Storages |
| NETWORK | |
|---|---|
Network Speed Baseline | 170 Gbps |
Inbound Traffic | 0 GB/month |
Outbound Traffic | 0 GB/month |
IPv4 | 0 |
Server Description
A bare-metal server featuring 192 dedicated Intel Xeon cores, eight G59 GPUs, and 1536 GB RAM for high-performance parallel computing.
Alibaba Cloud ecs.ebmgn9t.48xlarge is a high-density, GPU-accelerated bare-metal server built on the x86_64 architecture. It features 192 dedicated physical Intel Xeon 8575C CPU cores operating at 2.6 GHz with a 1:1 thread-to-core ratio. The instance is configured with 1536.0 GB of system memory, translating to 8.0 GB of RAM per core, and incorporates eight G59 GPUs with a total of 256 GB of VRAM. Network connectivity is established with a baseline bandwidth of 170 Gbps. A key architectural tradeoff is the absence of local storage, necessitating reliance on network-attached storage solutions. This server is engineered for demanding workloads such as large-scale machine learning, deep learning training, and GPU-accelerated scientific simulations.
Economics
Alternatives
Servers of the Same Family
| INSTANCE | vCPUs | MEMORY | GPUs |
|---|
Similar Servers
| INSTANCE | VENDOR | vCPUs | MEMORY | GPUs |
|---|