Written by Kristoffer Gulmark Poulsen
We are thrilled to announce that three new GPU (Graphics Processing Units) options are now available for CBS researchers on UCloud virtual machines: NVIDIA A10, A40, and A100.
GPUs are specialized hardware designed for parallel processing of data. Their ability to handle thousands of simple calculations simultaneously makes them highly efficient for computing tasks with machine learning, deep learning/AI and scientific simulations where massive parallelism accelerates performance significantly.
For those seeking high-performance options, you can initiate a dedicated virtual machine with the following specifications:
✔ 6x NVIDIA T4 GPUs with up to 210GB memory
✔ 4x NVIDIA A10 GPUs with up to 210GB memory
✔ 3x NVIDIA A40 GPUs with up to 420GB memory
✔ 2x NVIDIA A100 GPUs with up to 420GB memory
CBS is fortunate to have substantial resources at our disposal, and we cannot wait for CBS researchers to get their hands on these machines!
If you need more information, please join us for the upcoming webinar “Train your ML/AI Model on GPUs“.
Looking for tutorials and documentation?
If you have any questions, feel free to contact Kristoffer Gulmark Poulsen or send an e-mail to rdm@cbs.dk