Upcoming MGHPCC (Holyoke) Annual Power Shutdown [May 22, 2023 8:00 AM ET - May 28, 2023 8:00 PM ET]) [Timeline and info]

[More Information]
New England Research Cloud

Expanding our GPU Offerings

March 27, 2024

We are delighted to announce that we have recently expanding our GPU offerings by adding more NVIDIA A100 SXM4 40GB GPUs to both our NERC OpenStack and NERC OpenShift infrastructure (including NERC OpenShift Containers and NERC Red Hat OpenShift AI [RHOAI] JupyterLab environments). You can now leverage these new GPUs to enhance your computational requirements, further empowering your research and projects with increased power and efficiency.

For NERC OpenStack:
You can explore the newly added "gpu-su-a100sxm4" flavors and their variations, which are readily available for use when launching a VM on NERC OpenStack, as described in our OpenStack GPU offerings.

For NERC OpenShift:
You can specify multiple GPUs within the Pod spec, provided your NERC-OCP (OpenShift) quota has sufficient GPU quotas. You can then specify the new A100 GPU as described here.

If you prefer the NVIDIA V100 GPU device, you can specify that too, as explained here, which can be another cost-effective option for your GPU-based workloads.

For NERC OpenShift AI (RHOAI):
NERC RHOAI shares the same allocation quota as your NERC-OCP (OpenShift) resource allocation, which is requested using the NERC’s ColdFront Web Interface. When launching the workbench, you can select the required Number of GPUs for your JupyterLab environment, as described here.

What are NVIDIA A100 SXM4 40 GB GPUs?
The NVIDIA A100 SXM4 40 GB GPU is a high-performance computing accelerator tailored for parallel computational tasks, notably in artificial intelligence (AI), machine learning (ML), deep learning (DL), data analytics, scientific simulations, graphics-intensive applications, and beyond.

Save costs with alternative GPU flavors:
It's important to note that we also offer cheaper tier A100s, V100s and K80s GPU Offerings for NERC OpenStack and V100 based GPUs for NERC OpenShift. If your workloads do not require the high-end capabilities of A100 GPUs, switching to V100s and K80s can be a cost-effective solution without compromising performance.

How can I get customized A100 SXM4 GPUs not listed in the current flavors?
We also provide customized A100 SXM4 GPU-based flavors, which are not publicly listed on our NVIDIA A100 SXM4 40GB GPU Tiers list. These options are exclusively available for demanding projects and are subject to availability. To request access or discuss further, please fill out this form.

Feel free to contact us via email (help@nerc.mghpcc.org) or submit a new ticket to the NERC's Support Ticketing System if you have any questions or concerns.

We are committed to continuously improving our services and providing the tools you need to excel in your projects. We hope these newly added GPU flavors will open up new possibilities for your research work.

Tags:
Previous Post: