A Cluster for high-performance computing (High Performance Computing or HPC in English) is a set of nodes connected to each other through a high-speed communications network, which operates under software that acts as a single management system, responsible for distribute workloads between nodes, automatically and transparently to the user as if it were a single computer.
In order to respond to the demand for processing large volumes of information or the need for high computing power, Exacore offers the HYPATIA high-performance computing (HPC) service, an integrated solution that is supported by an 18-node supercomputer (16 CPU nodes and 2 GPU nodes), with the following main features:
*Front Node o SMS: 1 server Dell EMC PowerEdge R640. (Virtualized)
*Compute Nodes CPU:
10 servers Dell EMC PowerEdge R640.
1 server Dell EMC PowerEdge R640 (Función original IML)
1 server Dell EMC PowerEdge R6525 (1 TB RAM)
2 servers Dell EMC PowerEdge R6525 (512 GB RAM)
3 servers Dell EMC PowerEdge R6525 (256 GB RAM
*GPU Nodes: 2 servers Dell EMC PowerEdge R740 (3 NVIDIA RTX6000)
We have approximately 8.6 TB of memory distributed over an average of 470 Gb per node (some nodes have up to 565 Gb of RAM).
* 2 servers Dell EMC PowerEdge R740 – MDS.
* 2 servers Dell EMC PowerEdge R740 - OSS.
It provides 233 TB of research data storage, operating on the BeeGFS file system (which transparently spreads user data across multiple servers).
Low Latency Network: 1 Switch Mellanox Quantum HDR
Out-of-Band Management Network: 1 Switch Dell EMC S3148
Provisioning Network: 1 Switch Dell EMC S3048-ON
Hypatia, is installed on OpenHPC 2.3 (with a base system GNU/Linux type CentOS 8.4), the works are managed through Slurm and more than 300 software modules are available for use by researchers in scientific areas and related.
ExaCore has the Microsoft Azure public cloud, and more than 90 specialized servers for your workloads such as IoT, artificial intelligence and modeling.
Includes software setup, support, and basic training. Does not include installation of research software, except for institutional software, or training in the use and parameterization of software.
All references of virtual machines and services available in Microsoft Azure for the WestUS2 and EastUS geographies.
Discount on referrals D16aV4, D32aV4, D8aV4, E16aV4, E32aV4, E8aV4, E16aV3, E32aV3, E4aV3, E8aV3, M32ms, NC12, NC6, NC12sV2, NC6sV2, NV12, NV12sV3.
Check the characteristics of the machines here
Availability 24/7 from anywhere in the world, connection through the SSH protocol for Linux machines and RDP for supported Windows and Linux machines, with assigned Uniandes account.
ExaCore has a private cloud for the provisioning of large capacity virtual machines. The entire platform is focused on obtaining the best performance from the equipment and providing users with an infrastructure that has all the necessary resources for the most demanding projects. For this reason, the entire assembly of the cloud has been carried out using the suite of VMware products, obtaining the best benefit, support and efficiency from our infrastructure.
*CPU compute nodes: 9 Dell EMC PowerEdge R6525 servers (1 TB RAM)
*GPU compute nodes: 1 Dell EMC PowerEdge R740 servers (3 NVIDIA RTX6000)
We have approximately 10TB of memory distributed in an average of 1TB per node.
Provides 30 TB of research data storage.
It is configured on Linux and Windows operating systems, therefore it supports custom software, with or without a graphical interface.
* This service is exclusively for internal investigators