When conducting research using computing, it is possible for the program and data to outgrow the resources available on a desktop or laptop computer. This can cause issues such as insufficient memory, disk space, and slow computation times. While modern computers typically have up to 16GB of memory, a small number of CPU cores, and several TB of disk space, certain research problems may require more resources. In such cases, it can be beneficial to request access to an ACE-DS HPC cluster.
ACE-DS has established a high-performance computing cluster comprising of one master node and four data nodes, all of which are equipped with NVIDIA GPUs for graphic acceleration.
The data nodes are equipped with 32*2 cores of Intel(R) Xeon(R) Gold 6226R processors, which provide a total of 640GB of RAM per node, making 10GB/core. Additionally, each of the data nodes has 50TB of storage, all of which is GPU-enabled.
For moderate workloads, there are workstations available that feature an i9 processor, 64GB of RAM, and 10TB of storage.
The center also provides a significant number of desktop computers that users can utilize for their data science and artificial intelligence work.