Symantec Management Platform (Notification Server)

Long story short we are pulling an inventory of how many workstations we currently have deployed.  Our Asset Management team has a different program from Altiris that they use to audit machines.  The problem is that we have Altiris reporting 200 more PCs in the 'All Windows Workstations' collection than our Asset Management program.

Assuming our Asset Management program is correct do you all have any ideas of why Altiris would be reporting 200 more computers than we have?
Is All Windows Workstations not a good collection to use?

it shows all managed workstations. Do you purge or retire computers if they haven't checked in for awhile? What is your general 'source of truth', Active Directory? What tool does the Asset Management team use?

If Asset Management is using a different method of scanning (i.e. no a management agent), I would expect some deviation.

As Jim mentioned, "stale" objects will continue to be listed in the All Windows Workstations collection (including any machines which have been rebuilt with a new name).  The following will get you a list of all "Active" Windows Workstations:

SELECT acid._ResourceGuid
FROM Inv_AeX_AC_Identification acid
JOIN vResource vr
ON acid._ResourceGuid = vr.Guid
WHERE acid.[System Type] LIKE 'Win%'
AND DATEDIFF(dd, [Client Date], GETDATE()) < 30
AND vr.IsManaged = 1

In the Explicit Exclusions section of the collection, exclude the "All Windows Servers" collection.  This will give you all non-Server Windows computers which have sent basic inventory within 30 days.  This works well for us as we send basic inventory every day; if you send it less frequently then you may want to increase the 30 above to a higher number.

Currently we have a purge set if the computer doesn't report for 45 days.  Typically we consider Asset Management as the general source of truth and they use a program that runs an Audit on the machine every 15 days.  The program they use is called TrackIT.   

I have talked to the Asset Management guy and the problem might actually be on their end so we're testing it further.  Thanks for the help up to this point guys, I'll keep you up to date.