Autosuggest best practices

Question

Dear community members, 
 I was wondering whether there is an upper limit of TUs to be used when creating autosuggest dictionaries. 
 As there is a minimum of 10000 TUs required to create a dictionary, is there a maximum number of TUs which, when reached, would impact the performance when using the autosuggest dictionary? 
 Has anyone some experience or best practice to share on large autosuggest dictionaries? 
 Also, from a performance perspective, is it recommended to use a single, large autosuggest dictionary or multiple, smaller dictionaries? 
 Thank you for your support and advices. 
 Regards, 
 Laurent

Oana Nagy · Answer

Hello Laurent Pierret , 
 There are no actual physical limits to the size of a dictionary- but depending on your system specs/speeds, the bigger a particular resource is (TB/TM/Dictionary etc.) the longer it takes to search natuarally. 
 Unless you prefer a physical dictionary, you can instead leverage your TM directly- 
 Just go to Proct Settings Translation Memories- Search and enable TU Fragment- 
 
 When leveraging direct from the TM- there is no minimum amount of TU's needed- so if you are new to Studio, this option will benefit you immediately.

Oana Nagy · Answer

Hello Laurent Pierret , 
 Laurent Pierret said: What about one large dictionary vs multiple smaller ones? 
 I think this would come down to personal preference as im aware of some users who have created dictionaries (ASD's) in excess of 1 million TU's 
 As you are using GS and maybe the subject matter is varied- then you might prefer smaller ASD's. 
 Either way, as you will already be doing with your TM's and TB's, make sure you make regualr back up copies.

Trados Studio > 1. Trados Studio

Autosuggest best practices