There are several options and ways to optimise the performance within
UniversalPlantViewer (UPV). This document is intended to help find the
best solution for your specific needs. UPV by itself is able to handle models and sessions of literally any
size, but the performance depends on the setup of the device in use.
Depending on the OS (Windows, Android, iOS), the CPU, GPU and RAM, there
can be restrictions and limits regarding the usage of UPV or any other
software. In this case, we recommend optimising your model with regard
to the following: Even though the amount of attributes play a minor role for the
overall model size, the number of attributes in the model can affect the
overall performance. There are several leverage points to improve
performance: Avoid unnecessary attributes Many authoring systems create
attributes that are of no informational value to the user, e.g., color,
shading, and other information relevant to the authoring system only. By
removing these attributes from the session in UPV Builder, you can
streamline the information for your session. The less attributes the
better the orientation in UPV and the faster the search and other
functions perform. Fixed attributes before custom attributes If you do not need to
change attribute values during the UPV runtime, preferrably use the SDK
import functionalities (e.g., ExcelImport plugin) before Custom
Attributes. Custom Attributes are loaded on model opening and extend
loading time noticeably if they are used extensively. Optimise attribute hierarchy In UPV you can define your custom
hierarchy, enabling you to streamline the structure of your data, thus
improving performance for the model and display speed of attributes at
object selection change. Especially for mobile devices with restricted memory management, the
scope of the model can have an impact on the performance. For these
mobile devices, we recommend using the package functionality in UPV
builder to reduce the full model to smaller packages, still enabling the
mobile user to view the different packages separately on their mobile
device. In the UPV settings there are several options to tweak the
performance of the viewer, influencing loading, caching and shading of
the model: Render quality level Low - display the model without shading.
Recommended for best performance Medium - display simple shading.
Recommended for overlapping surfaces High - best quality including
shadows. Highest ressource comsumption Triangles limit Defines the number of triangles displayed
simultaneously. The more triangles are being displayed, the better the
view gets but the more performance is needed. Memory mesh cash size (triangles) These settings define the
number of triangles kept in the memory. This has influence on the RAM
usage. Setting the memory mesh cash size to 0 disables the caching -
this results in slightly longer reloading times but lowers the memory
usage of the viewing device. Memory model cache size (bytes) Sets the maximum amount of RAM
available to the model. This setting controls the memory usage and can
be adapted to suit devices with lower memory sizes. We recommend to adapt these settings/options to values that suit
either most of your devices or - for maximum compatibility - to the
lowest common setting. Users with more powerful workstations and devices
with regard to CPU, GPU and RAM can adapt these values individually and
save their settings to be loaded manually subsequently. Alternatively or if none of the above measures leads to the expected
results, you can use our browser-based viewing technology. This unique
technology enables you to view your model in any industry standard web
browser without any installation, plugins, or other requirements except
for an internet (or intranet) connection to the server. This technology
relocates the computing into the cloud or your corporate intranet. This
requires powerful hardware on the server side but therefore enables
literally any device to view models of any size at top performance.
Appendix - Performance
Optimisation
Best Practise for Optimal
Performance
Optimising the Model and
Data
Attribute Information
Model Scope
Performance Options
The Alternatives
Upgrade the Hardware
One
alternative would be to upgrade the hardware on your devices (CPU, GPU,
RAM on workstations or desktops or better equipped mobile devices). This
often leads to a much better performance on the devices in question.
When acquiring new hardware, consider these parameters for optimum
performance.Browser-Based Viewing
