News on NEMO2 Procurement

News on NEMO2 Procurement

Published: 12 Dec 2023 by NEMO Team

UPDATE 14.02.2024: Ethernet hardware updated.

Dear NEMO users,

Over the past few months, a tender for the new NEMO2 hardware has been conducted. While negotiations are ongoing, the parties involved are not allowed to talk about their business in public. Therefore, there has been no news from NEMO as NEMO1 is still in operation and nominally working.

Negotiations have now been concluded and the first CPU partition of NEMO2 has been ordered. The first partition of the HPC system will consist of approximately 140 Milan nodes with 128 cores, 512 gigabytes of RAM, 100 gigabit Ethernet, 100 gigabit High Speed OmniPath and two terabytes of local NVMe flash memory. The high-speed parallel storage provides 1000 terabytes (one petabyte) of usable capacity on full flash NVMe, a throughput of around 90 gigabytes/s write and over 200 gigabytes/s read (extrapolated IOR benchmark values).

The initial investment has been optimized for low total cost of ownership (TCO) over the expected lifetime of the systems, with significant consideration given to energy costs. From NEMO2 onwards, a new operating model will be used and new services and functions will be introduced.

A tender for an additional GPU partition will be opened in early 2024. If you have specific requirements for GPU architectures, please let us know your requirements and benchmark suggestions in advance.

There will also be a tender for a second CPU partition with an even stronger focus on energy efficiency.

NEMO2 Procurement CPU Partition Storage Milan Weka

Latest Posts

NEMO2 Production Mode

NEMO2 has officially launched, transitioning from a testing phase to full production with expanded hardware, including AMD Instinct MI300A and Nvidia L40A nodes. NEMO1 is being phased out, with limited resources available until May 31st. Users are encouraged to transition to NEMO2 and consult the wiki for details.

Read more

Published: 14 May 2025

NEMO2 Conda

NEMO2 uses Miniforge for conda environments, offering a streamlined setup with conda-forge as the default repository. The Miniforge module auto-initializes conda, simplifying environment activation without modifying shell profiles.

Read more

Published: 14 May 2025

Genoa Nodes Delivered

The AMD Genoa, Machine Learning and AI partitions for NEMO2 were delivered on December 4th. The acceptance of the storage has been delayed, so that NEMO2 could not yet start this year. However, calculations with the Milan nodes in NEMO1 are still possible.

Read more