# Server sizing guide question **URL:** \ **Category:** Ignition\ **Created:** [October 24, 2024, 12:20pm UTC](https://forum.inductiveautomation.com/t/server-sizing-guide-question/97725 "2024-10-24T12:20:30Z")\ **Posts on this page:** 2\ **Page:** 1
**Author:** ![Pdeswert](https://avatars.discourse-cdn.com/v4/letter/p/a8b319/32.png) [@Pdeswert](https://forum.inductiveautomation.com/u/Pdeswert)\ **Post date:** [October 24, 2024, 12:20pm UTC](https://forum.inductiveautomation.com/t/server-sizing-guide-question/97725/1 "2024-10-24T12:20:31Z")
The sizing guide only talks about servers that have both tag processing and visualisation. However I was wondering if the requirements differ much when you split the servers into a back and front end. ---
**Author:** ![pturmel](https://sea1.discourse-cdn.com/inductiveautomation/user_avatar/forum.inductiveautomation.com/pturmel/32/3295_2.png) [@pturmel](https://forum.inductiveautomation.com/u/pturmel)\ **Post date:** [October 24, 2024, 12:32pm UTC](https://forum.inductiveautomation.com/t/server-sizing-guide-question/97725/2 "2024-10-24T12:32:59Z")
Did you see this forum topic? > [@Microsoft Virtual Machine Suggestion](https://forum.inductiveautomation.com/t/microsoft-virtual-machine-suggestion/86787): > > All, Does anyone have a suggestion on what size of VM should one purchase? I'm currently using B2MS. When I try to view historical data. It bogs down and restarts. Thanks, C I/O-only gateways are relatively light loads. You can approximate the workload of remote gateway connections (remote tag providers, remote history) as similar to Vision client connections, even when the front end is Perspective. Visualization-only gateways only avoid the historization workload, so probably will need near the full forecast capacity. As noted in the sizing guide and that topic, actual workloads vary greatly--you should test, test, test.