LUMI Overview¶
LUMI is one of the three European pre-exascale supercomputers. It's an HPE Cray EX supercomputer consisting of several partitions targeted for different use cases. The largest partition of the system is the "LUMI-G" partition consisting of GPU accelerated nodes using a future-generation AMD Instinct GPUs. In addition to this, there is a smaller CPU-only partition, "LUMI-C" that features AMD EPYC "Milan" CPUs and an auxiliary partition for data analytics with large memory nodes and some GPUs for data visualization. Besides partitions dedicated to computation, LUMI also offer several storage partitions for a total of 117 PB of storage space.
LUMI-C: The CPU Partition¶
The LUMI-C partition consists of 1536 compute nodes with an estimated combined LINPACK performance of ca. 8 Petaflops. Each LUMI-C compute nodes are equipped with 2 AMD EPYC 7763 CPUs with 64 cores each running at 2.5 GHz for a total of 128 cores per node. The cores have support for 2-way simultaneous multithreading (SMT) allowing for up to 256 threads per node.
LUMI-D: The Data Analytics Partition¶
LUMI-D is intended for interactive data analytics and visualization. It is also a good place run pre- and post-processing jobs that require a lot of memory. It consists of a 8 nodes with large memory capacity (4 TB) and 8 nodes with NVIDIA A40 GPUs. Each LUMI-D compute nodes are equipped with 2 AMD EPYC 7742 CPUs with 64 cores each running at 2.25 GHz for a total of 128 cores per node.
LUMI-P and F: Parallel Filesystems¶
LUMI has two Lustre parallel file systems consisting of:
A main storage partition (LUMI-P) composed of 4 independent Lustre file systems with an aggregated performance of 240 GB/s and a 20 PB storage capacity each
A flash storage partition (LUMI-F) optimized to support high IOPS rates with an aggregated performance of 1740 GB/s and 7 PB of storage capacity
LUMI-O: The Object Storage¶
Object storage is a data storage architecture that manages data as objects instead of a file hierarchy. Each object includes the data, the metadata and a globally unique identifier. This partition may be used for storing, sharing and staging your data. It's based on Ceph and has a storage capacity of 30 PB.