Echelon is exceptionally fast. By running all key computational tasks on GPUs it achieves unprecedented speed. The full SPE10 benchmark is completed in under one minute and Echelon is typically 10x to 50x faster than leading commercial simulators.
Echelon scales efficiently to very large systems. It runs across multiple GPUs housed in multiple nodes. Cases with up to 220 million cells have been run in less that 40 minutes with 48 GPUs on an 8 node cluster.
Echelon runs on superfast NVIDIA GPUs in a workstation or cluster. It can handle up to 40M cells on a workstation or 80M cells in a single cluster node. Affordable, compact, high density supercomputing.