System recommendations for GeoDict
The GeoDict software by Math2Market GmbH is available for:
- Windows 7, 8, and 10 (64 bit)
- Linux (64 bit) with at least glibc 2.12
Math2Market GmbH guarantees that the current GeoDict release is always compatible with the following operating systems:
- Windows 10 Home / Pro (64 bit) for Current Branch for Business (CBB) (to Math2Market GmbH Ready-for-Windows page)
Parallel and efficient
Even if GeoDict' solvers sequential computations are remarkably fast, many of them can be run even faster using parallelization.
To parallelize processes, GeoDict takes advantage of multiple cores on all platforms using the Message Passing Interface (MPI) protocol, for which the MPI library must be installed. Click the links for information on downloading and installing MPI on Windows or Linux systems.
The best performance is achieved by GeoDict's high-performance version for 64-bit Linux-parallel computers.
Modelling and simulation of ultralarge-scale, highly representative models can be efficiently carried out using multi-core processors.
Work with GeoDict on your workstation PC, a shared memory workstation, or on high-performance computing clusters.
The right configuration for you
The size of the structure models that GeoDict handles depends on the size of available RAM.
The budget-based configurations below are recommendations in consultation with our hardware partner Delta Computer.
We do not sell the hardware.
Contact us for help to find the optimal GeoDict system configuration for you
or for any questions about cluster-installations.
|Budget €||Cores||RAM GB||Graphics RAM GB||Visualization in Voxel||Simulations in Voxel|
|~ 1.500||4 (1x 4)||32 DDR4||1 GB||~ 1 billion||~ 0.5 billion|
|~ 7.000||16 (2x 8)||256 DDR4||2 GB||~ 9 billions||~ 3 billions|
|~ 18.000||36 (2x18)||512 DDR4||2 GB||~ 17 billions||~ 6 billions|
|~ 30.000||40 (4x10)||1024 DDR4||2 GB||~ 34 billions||~ 12 billions|
|~ 72.000||60 (4x15)||3072 DDR4||2 GB||~ 103 billions||~ 36 billions|
|On request||Cluster||Cluster||-||32 Mio / GB||12 Mio / GB|