• Number of Nodes: 34
• CPUs:
  • 2 x AMD EPYC 9754 with 128 cores per CPU = 256 cores per node
  • Base Frequency 2.25 GHz. Up to 3.1 GHz
  • Total of 8,704 cores
• Memory:
  • 768 GB of RAM per node
  • Total of 25.5 TB of RAM
• Network:
  • Infiniband at 400 Gbps
• Storage:
  • 4 TB SSD

• Number of Nodes: 160
• CPUs:
  • 2 x AMD EPYC 7H12 with 64 cores per CPU = 128 cores per node
  • Base Frequency 2.60 GHz. Up to 3.3 GHz
  • Total of 20,480 cores
• Memory:
  • 512 GB of RAM per node
  • Total of 80 TB of RAM
• Network:
  • Infiniband at 100 Gbps
• Storage:
  • 1 TB SSD

• Number of Nodes: 4
• CPUs:
  • 2 x AMD EPYC 7742 with 64 cores per CPU = 128 cores per node
  • Base Frequency of 2.25 GHz. Up to 3.4 GHz.
  • Total of 512 cores
• Memory:
  • 1 TB of RAM per node
  • Total of 4 TB of RAM
• Network:
  • Infiniband at 200 Gbps with the rest of the cluster
  • Infiniband at 8 x 200 Gbps internally to the 4 nodes
• Storage:
  • 1 TB SSD
• GPUs:
  • 8 x Nvidia A100 with 40 GB VRAM on each node
  • Total of 32 Nvidia A100
  • Total of 110,592 CUDA cores
  • Total of 13,504 Tensor cores
  • Total of 1280 GB of VRAM

• Number of Nodes: 24
• CPUs:
  • 2 x AMD EPYC 7H12 with 64 cores per CPU = 128 cores per node
  • Base Frequency of 2.60 GHz. Up to 3.3 GHz
  • Total of 3,072 cores
• Memory:
  • 2 TB of RAM per node
  • Total of 48 TB of RAM
• Network:
  • Infiniband at 200 Gbps
• Storage:
  • 5 TB SSD

• Number of Nodes: 126
• CPUs:
  • 2 x Intel Xeon Gold 6230R with 26 cores per CPU = 52 cores per node
  • Base Frequency of 2.10 GHz. Up to 4.00 GHz
  • Total of 6,550 cores
• Memory:
  • 192 GB of RAM per node
  • Total of 23.5 TB of RAM
• Network:
  • Infiniband at 100 Gbps
• Storage:
  • 1 TB SSD

All our critical servers run on a high-availability VMware VSAN cluster and Vmotion. This cluster consists of:

• 5 x Lenovo at 2.66 GHz and 512 GB of RAM each
• 40 TB of shared VSAN storage
• Internal network of 25 Gbps
• Network with the rest of the cluster at 10 Gbps

All processing nodes store user data in two shared storage clusters:

• FSCRATCH of 329 TB
  • 329 TB of state-of-the-art NVME SSDs
• GPFS of 555 TB
  • Layer of 55 TB SSDs
  • Layer of 500 TB NL-SAS HDDs
  • Infiniband connection with the rest of the cluster

Object Storage of 6 PB

At end 2021, the infrastructure of Picasso was modified, adding space for additional racks, improvements in the air conditioing, power facilities and UPS. That allowed the upgrade of almost all the nodes, an upgrade that has moved Picasso to a new level, with more than 30.000 compute cores and a pool of machine learning optimized machines with 32 dedicated A100 GPUs.

•  2 x AMD EPYC 7H12 processors/node x 64 cores = 128 cores/node
•  160 nodes x 128 cores = 20.480 cores
•  2.60 GHz per core (base). Up to 3.3 GHz
•  160 nodes x 512 GB RAM = 80 TB total RAM
•  Infiniband HDR network at 100 Gbit/s
•  Localscratch of 950 GB

•  2 x AMD EPYC 7742 processors/node x 64 cores = 128 cores/node
•  4 nodes x 128 cores = 512 cores
•  2.25 Ghz per core (base). Up to 3.4 GHz
•  4 nodes x 1024 GB RAM = 4 TB total RAM
•  4 nodes x 8 Infiniband HDR network at 200 Gbit/s (only inside DGX cluster)
•  Infiniband HDR network at 200 Gbit/s (all cluster)
•  4 nodes x 8 Nvidia A100 GPUS = 32 Nvidia A100 GPUs
•  32 Nvidia A100 GPUs x 3456 cores CUDA FP64 = 110.592 cores CUDA on GPUs.
•  32 Nvidia A100 GPUs x 422 Tensor cores for AI =  13504 Tensor cores on GPUs.
•  32 Nvidia A100 GPUs x 40 GB VRAM per GPU = 1280 GB VRAM on GPUs.
•  Localscratch of 950 GB

• 2 x Intel Xeon Gold 6230R processors/node x 26 cores = 52 cores/node
• 126 nodes x 52 cores = 6552 cores
• 2.10 GHz per core. Up to 4 GHz
• 126 nodes x 192 GB RAM = ~ 24 TB total RAM
• Infiniband HDR network at 100 Gbit/s
• Localscratch of 950 GB

• 2 x AMD EPYC 7H12 processors x 64 cores = 128 cores
• 24 nodes x 128 cores = 3072 cores
• 2.60 GHz per core. Up to 3.3 GHz
• 24 nodes x 2 TB RAM = 48 TB total RAM
• Infiniband HDR network at 200 Gbit/s
• Localscratch of 3.5 TB

All our critical virtual desktops and servers are hosted on a VMware VSAN cluster with High Availability and VMotion. This cluster consists of:

•  5 x Lenovo at 2.66 Ghz with 512 GB of RAM each
•  40 TB shared VSAN storage.
•  25 Gbits internal network.
•  10 Gbits external network.

At mid 2018, the shared storage system has been upgraded, adding a new high performance filesystem with 555 TB and a long term object storage with a raw capacity of 6 PB. This new upgrade increases the performance of the filesystem thanks to a super fast SSD layer of 55TB. Hot data is kept on the faster storage layer, being moved down to a second level of 500 TB NL-SAS disks when data is not used. The storage uses the infiniband network between all nodes of the system.

•     4016 compute cores
•     74 TFLOPS in compute nodes + 33 TFLOPS in GPUs

• 48 HP-SL230G8 x 2 processors x 160 GLOPS = 16 TFLOPS
• 32 GPUS M2075 x 1030 GFLOPS = 33 TFLOPS
• 48 x 2 E5-2670 processors x 8 cores = 768 cores
• 2.60GHz per core
• 48 * 64 GB RAM = 3 TB total RAM
• Infiniband FDR network at 54.54 Gbit/s
• Shared scratch in lustre filesystem

• 7 DL980G7 x 8 processors x 96 GLOPS = 5 TFLOPS
• 7 x 8 E7-4870 processors x 10 cores = 560 cores
• 2.40GHz per core
• 7 x 2 TB RAM = 14 TB total RAM
• 7 x 2 TB RAID5 = 14 TB local scratch
• Infiniband QDR network at 32 Gbit/s

• 168 IBM dx360 M4 x 2 processors x 160 GFLOPS = 53 TFLOPS
• 168 x 2 Intel E5-2670 processors x 8 cores = 2668 cores
• 2.60GHz per core
• 168 * 32 GB RAM = 5.4 TB total RAM
• Infiniband FDR network at 40 Gbit/s
• Shared scratch in lustre filesystem

All our critical virtual desktops and servers are hosted on a VMware ESX Enterprise cluster with High Availability and VMotion. This cluster consists of:

•  2 x HP-DL380G5 at 2.66 Ghz with 32 GB of RAM each

For those virtual machines and user grade desktops that doesn’t need High Availability support, there is set of machines with the ESXi hypervisor installed. These servers are:

•  3 x HP-DL380G5 at 2.66 Ghz with 16 GB of RAM each
•  4 x HP-DL385G7 at 2.13 Ghz with 124 GB of RAM each

At mid 2017, the cluster of 41 nodes with AMD Opteron 6176 has been replaced with 168 IBM dx360 M4 compute nodes. This new upgrade increases the amount of available compute cores up to a total of 4016. Now, all nodes share the same infiniband network and a common architecture, and are unified behind an unique Slurm queue system that delivers jobs to the most adecuate equipment based on requested resources.

•       4016 compute cores
•       74 TFLOPS in compute nodes + 33 TFLOPS in GPUs
•       750 Tb of shared storage with lustre
•       Infiniband QDR/FDR interconnection network.

• 48 HP-SL230G8 x 2 processors x 160 GLOPS = 16 TFLOPS
• 32 GPUS M2075 x 1030 GFLOPS = 33 TFLOPS
• 48 x 2 E5-2670 processors x 8 cores = 768 cores
• 2.60GHz per core
• 48 * 64 GB RAM = 3 TB total RAM
• Infiniband FDR network at 54.54 Gbit/s
• Shared scratch in lustre filesystem

• 7 DL980G7 x 8 processors x 96 GLOPS = 5 TFLOPS
• 7 x 8 E7-4870 processors x 10 cores = 560 cores
• 2.40GHz per core
• 7 x 2 TB RAM = 14 TB total RAM
• 7 x 2 TB RAID5 = 14 TB local scratch
• Infiniband QDR network at 32 Gbit/s

• 168 IBM dx360 M4 x 2 processors x 160 GFLOPS = 53 TFLOPS
• 168 x 2 Intel E5-2670 processors x 8 cores = 2668 cores
• 2.60GHz per core
• 168 * 32 GB RAM = 5.4 TB total RAM
• Infiniband FDR network at 40 Gbit/s
• Shared scratch in lustre filesystem

All our critical virtual desktops and servers are hosted on a VMware ESX Enterprise cluster with High Availability and VMotion. This cluster consists of:

•  2 x HP-DL380G5 at 2.66 Ghz with 32 GB of RAM each

For those virtual machines and user grade desktops that doesn’t need High Availability support, there is set of machines with the ESXi hypervisor installed. These servers are:

•  3 x HP-DL380G5 at 2.66 Ghz with 16 GB of RAM each
•  4 x HP-DL385G7 at 2.13 Ghz with 124 GB of RAM each

At the begining of 2013, all supercomputing resources has been upgraded. Now, they share a common architecture, and are unified behind an unique Slurm queue system that delivers jobs to the most adecuate equipment based on requested resources.

• 48 HP-SL230G8 x 2 processors x 160 GLOPS = 16 TFLOPS
• 32 GPUS M2075 x 1030 GFLOPS = 33 TFLOPS
• 48 x 2 E5-2670 processors x 8 cores = 768 cores
• 2.60GHz per core
• 48 * 64 GB RAM = 3 TB total RAM
• Infiniband FDR network at 54.54 Gbit/s
• Shared scratch in lustre filesystem

• 7 DL980G7 x 8 processors x 96 GLOPS = 5 TFLOPS
• 7 x 8 E7-4870 processors x 10 cores = 560 cores
• 2.40GHz per core
• 7 x 2 TB RAM = 14 TB total RAM
• 7 x 2 TB RAID5 = 14 TB local scratch
• Infiniband QDR network at 32 Gbit/s

•  41 HP-DL165G7 x 2 processors x 110 GFLOPS = 9 TFLOPS
•  41 x 2 Opteron 6176 processors x 12 cores = 984 cores
•  2.30GHz per core
•  41 x 96 GB RAM =  4 TB RAM
•  Gigabit ethernet network.
•  41 * 6 TB RAID5 = 246 TB total scratch

All our critical virtual desktops and servers are hosted on a VMware ESX Enterprise cluster with High Availability and VMotion. This cluster consists of:

•  2 x HP-DL380G5 at 2.66 Ghz with 32 GB of RAM each

For those virtual machines and user grade desktops that doesn’t need High Availability support, there is set of machines with the ESXi hypervisor installed. These servers are:

•  3 x HP-DL380G5 at 2.66 Ghz with 16 GB of RAM each
•  4 x HP-DL385G7 at 2.13 Ghz with 124 GB of RAM each

The year 2007 supossed a great improvement for UMA supercomputing facilities. A new building dedicated to excellence in research was finished, and all supercomputing equipment was moved there. It counts with two DataCenter rooms, one for high performance machines, and the other one for equipment with less cooling requirements, like tape libraries.

At this time, there were three different architectures inside the same room.

Two racks were used to hold only cpus, memory, and all interconnection cables necessary to reach all the shared memory from any compute core. A third rack was used for operating system and additional scratch storage.

Program parallelization in this kind of supercomputers (shared memory ones), is usually easier than on a cluster, since it appears as a single computer to end users.

Pablo is a cluster with 512 POWER970 microprocessors and 1TB of RAM. It is composed by 256 blades equipped with two POWER cpus and 4 GB of RAM.

This cluster belongs to the RES (Red Española de Supercomputación – Spanish Supercomputing Network), and its time is shared between RES users and University of Málaga.

Three racks are used to host all blades, and another one for administration servers and storage infrastructure.
All blades are interconnected by Mirynet, a fast and with a very low latency network very suitable for clusters.

There was available a x86 cluster with 80 cores and 160GB of RAM, composed by 10 blades with two Xeon quadcore processors (8 cores per blade) and 16 GB of RAM.

These blades are interconnected by an InfiniBand network which provides a very high intercommunication speed that improves parallel processes.

This cluster is still in use for backend support.

It was on 1997 when the first supercomputer was installed in the University of Málaga, dedicated exclusively to the execution of scientific applications by the UMA researchers. It was called “Picasso”, a SGI Origin 2000 machine with 16 processors at 200Mhz and 4 GB of RAM that gave an uninterrupted service for over 10 years.

It now enjoys a deserved retirement in the UMA museum of technology.