Client System Requirements
Operating systems | 64-bit Supported Windows Server 2008 R2 or above Windows 7 or above |
|---|---|
CPU* | 4 Core |
Memory* | 4 GB (Minimum),8GB (Recommended) of Java heap size is enough for most evaluation purposes |
Hard Disk | 50GB hard disk space (or more to keep logs in text files if required) |
CPU
2.8GHz or more suggested. Should monitor CostOS CPU usage and allocate more Cores if required. This is because each created project can create new workflows, new custom Columns, complex Calculations, etc.
Memory
Bim Module-Plugin Requires GPU for 3D rendering.If gpu is not installed on the machine the performance rate may fall down but, in any case, CostOS client is optimized for this kind of rendering.
Database Network Connection
If you are using a different Machine as a Database server must be on the same physical network for optimal performance without network latency.
Read More
BIM Module-Plugin included?
Before Purchasing a VGPU make sure that Windows Server Drivers for OpenGL 4.6 are available. Lots of Server Side GPUs are designed for Machine Learning and do not provide 3D Visualization Drivers.
You can run CostOS client into a virtual machine. On that case we need to consider and design the hardware requirements depending the number of concurrent users and the requirements above.
BIM Module-Plugin included?
Before Purchasing a VGPU make sure that Windows Server Drivers for OpenGL 4.6 are available. Lots of Server Side GPUs are designed for Machine Learning and do not provide 3D Visualization Drivers.
Supported web browsers
You should use a browser that supports HTML5 content. Any latest release of Mozilla Firefox or Google Chrome will do.Internet Explorer is not supported. Microsoft Edge will work but will not visualize BIM content.
Running CostOS Workstation on a Shared Virtual Machine
When multiple users run CostOS Workstation on the same virtual machine simultaneously, the hardware requirements above must be scaled according to the number of concurrent users. Each CostOS instance runs as an independent process and does not share memory with other instances, so resources must be allocated per user.
As a general rule, multiply the minimum requirements by the number of users who will be logged in and actively working at the same time:
Concurrent Users | CPU (Cores) | Memory (Minimum) | Memory (Recommended) |
|---|---|---|---|
1 user | 4 cores | 4 GB | 8 GB |
2 users | 8 cores | 8 GB | 16 GB |
4 users | 16 cores | 16 GB | 32 GB |
8 users | 32 cores | 32 GB | 64 GB |
These figures represent peak load. If users are unlikely to all perform intensive operations simultaneously, CPU core count can be moderately reduced, but memory should always be allocated in full as each instance reserves its own dedicated heap.
If the BIM module is in use, a vGPU supporting OpenGL 4.6 is required. Note that many server-grade GPUs are designed for machine learning workloads and do not include 3D visualisation drivers — confirm OpenGL 4.6 support before purchasing. The GPU capacity must also be sufficient to handle the number of concurrent BIM sessions expected on the machine.
For large deployments with many concurrent users, Nomitech recommends spreading users across multiple virtual machines rather than increasing the size of a single machine, as this improves stability and makes it easier to manage performance and availability independently.