Monster PDF's over several hundred MB cause huge problems over a 100M/B network. You'll find that the PC swap files are usually not big enough to handle them.
A way to reduce file size, besides dividing them in two, is to copy the file to a folder on a local drive, then open it in Acrobat, and Save As to a JPEG2000 (.JPF) file. Once that's done, close the original file, and Create a new PDF from the collection of JPF's that were created. There will be a JPF for each page of the original. Then save the new PDF. It will be 4 to 12 time smaller than the original. if the file included OCR-able text, run the OCR Text capture and save another 10% on file size when you save it.
If the problem involve large engineering drawings, you may need to change the orientation so they're very tall rather than very wide before you can do the JPF feature.
Another reason to compress the PDF's for network storage is that when a user tries to open a monster file, it clogs the network for everyone else.
Another solution you might want to consider is a Document Management system. Laserfiche stores image files as multi-page TIFF's rather than PDF's, and serves the pages across the network individually, rather than forcing the workstation to open an entire file. The indexing for the document is kept in a master database, so you can still search for a word, and be fed the pages where it occurs from the list of hits.
As you can tell, we've had lots of experience with this issue, and with solutions to it!
Fred Wagner