Internal Facilities

  Home / Research / Internal Facilities
                               

Computing Resources

CPU Clusters

Venus

As a single-node computer, it has 16 Intel Xeon E7-8860v3 CPUs (a total of 256 processors) which can run 512 threads in parallel, 4TB shared memory and a total of 180TB storage. Please contact Prof. Yi Mao (ymao@tsinghua.edu.cn) to open an account.                                                                        

Orion

As a powerful Dell HPC cluster, it has 80 nodes, with a total of 2240 processors, and 480TB storage. Each node has two Intel Xeon Gold 6132 CPUs and 96GB memory. Please contact Prof. Yi Mao (ymao@tsinghua.edu.cn) to open an account.                                                                            

Sirius

As a powerful Dell computing and storage cluster, it consists of 12 computing nodes with a total of 1152 cores (each node with 4 Intel Xeon Gold 6252 2.1GHz CPU, totally 96 cores and 2TB memory per node), a total of 92 TB SSD storage and 1.3 PB HDD storage. Please contact Prof. Yi Mao (ymao@tsinghua.edu.cn) to open an account.

Jupiter

As a Dell storage cluster, it has 840 TB hard disk storage, for saving permanent data. Contact: Prof. Yi Mao (ymao@tsinghua.edu.cn).                                                                            
                                                                           
                                                                           
                                                                           
                                     

GPU Clusters

Metis

As a Dell GPU cluster, Metis has two intel Xeon Gold 6248 CPUs (a total of 40 cores), 784 GB shared memory, 12.48TB SSD + 168TB hard disk storage, and 10 nvidia v100 Graphics cards (each with 32GB memory, PCI-e). Please contact Prof. Yi Mao (ymao@tsinghua.edu.cn) to open an account.                                                                            
                                                                           
   

R2D2

This is a miniature desktop workstation optimized for machine learning, equipped with a GTX 1080Ti Graphics Card featuring 11GB GDDR5X 352-bit memory, 3584 CUDA cores, and 11GB Frame Buffer. Please contact Xiangyun Long (longxy15@mails.tsinghua.edu.cn) to open an account.                                                                            
                                                                           
  

Observing Resources

TNT

DoA has a 50% share on the 80cm Tsinghua-NAOC Telescope (TNT).                                                                            
                                                                             
                                                                             
                                                                             
                                                                           
   

TAP

DoA faculty, postdocs and students can apply for telescope time, currently on CFHT, Palomar 5m and LCO 1m telescopes. More detailed information can be found at http://info.bao.ac.cn/tap/