Application as early-adopter of GPGPU cluster Alex

FAU purchased “Alex” as Tier3 and NHR GPGPU cluster with Nvidia A100 (very good double precision floating point performance) and A40 (extraordinary single precision floating point performance) GPGPUs. The Tier3 part serves the basic needs of FAU (and RRZE’s partner universities). The NHR part is accessible by researchers from universities all over Germany after approval of a scientific project proposal. If you are from FAU and your demands exceed the free basic tier, you also have to apply for an NHR project.

The hardware of Alex will be delivered in the 2nd half of November and we are now looking for experienced HPC users as early adopters who will help us to bring Alex into stable production. Early adopter have to work closely with the HPC team to optimize system configuration and centrally installed software. As a reward, early adopters get plenty of compute cycles, but they have to expect configuration changes and stability issues. The initial early adopter phase is expected to run from early December until early February.

Please fill in the following form if you are an experienced HPC user and interested to become an early adopter. All registrations are personal ones; multiple people from one workgroup may apply.

The selection of proposals will be based on the expected outcome and diversity of groups / applications / scientific domains. Projects which can use significant GPU hours efficiently will be preferred.