Services and capacities
Supercomputer (HPC)

Our HPC service is focused on solving situations of high impact simulations that occur in science and industry. The infrastructure that we offer together with our qualified personnel , results in a better use of the architecture, and shorter time of processing use.

An additional value to our service is the timely advice of scientists, and experts in HPC issues, who will help us find solutions and improvements to the problems. We offer the following options:


SIMULATION
Focused to show results based on program outputs to represent the data in a graphical way.

ANALYSIS
Mechanism of validation of data and improvements in the processes of compilation, execution and use of codes.

Bigdata
Analysis of large blocks of information with infinity of variables to find their behavior.

Our supercomputer
Thubat Kaal 2.0
Bull Equipment: 257.24 Theoretical Teraflops

Computational capacity of 257,243 TERAFLOPS in double precision (182.04 Intel Skylake Xeon x86 64 Bit, 75.2 NVIDIA GPU P100). It is composed of 86 calculation nodes of which 82 nodes are of INTEL XEON technology X86 BITS SKYLAKE (Purley Platform) and 4 nodes contain 4 NVIDIA P100 plates each. It has a High Availability Luster file system with 1.8 PB available. GREEN TECHNOLOGY and friendly to the environment. GREEN TECHNOLOGY and friendly to the environment.

Hardware
4 RACK 42 U with PDUs and compatible power cables for the equipment. 86 expandable calculation nodes, with the following characteristics:

64 compute nodes BULL R440-E5
  • 2 Intel Xeon Skylake 6130 16c processors, 2.1 - 3.7 ghz (32 CORES per node)
  • 192 GB RAM (6 GB per CORE)
18 compute FAT nodes BULL R440-E5
  • 2 Intel Xeon Skylake 6130 16c processors, 2.1 - 3.7 ghz (32 CORES per node)
  • 384 GB RAM (12 GB per CORE)
4 GPU nodes BULLX R421-E4K
  • 2 processors INTEL XEON E5-2680 14c, 2.4 GHz (28 CORES per node)
  • 128 GB RAM (4.57 GB per core)
  • 4 NVIDIA P100 cards per node.
2 management nodes for services configured in high availability cluster mode (HA Active-Active)

2 access nodes for user login services and home space configured in active active high availability (HA Active Active) cluster mode for login in Active-Passive mode (HA-Active-Passive) for space in Home. Directory space / home 200 TB.

High performance Infiniban EDR 100GBPS network configured in an All to All topology.
They make us the cluster with the fastest infinity and currently in Mexico
10 Gbps Ethernet management

File system Luster for Scratch of 1.8 PB or 1843 Tb approx.

Software
Operating system: RedHat 7.3 licensed in the entire cluster.
System scheduler: Slurm.