I think the fastest and safest way to do this can be elaborated in below steps
* Identify all rows (primary key) that will get affected and store it in excel sheet be executing below query
SELECT A.UID, B.GrossAmount, B.Units, B.GrossAmount
from TUnit A
join TUnit B
on A.UID = B.UID
and A.TID = B.TID
and A.AID = B.AID
and A.TAID = B.TAID
join #TmpTUnit U
on A.UID = U.UID
and A.TID = U.TID
and A.AID = U.AID
and A.TAID = U.TAID
and A.TType = U.TeType
* transform your update query into single primary key based query. Since you now have 100K rows in the excel sheet you can build the update query very easily.
UPDATE TUNIT set RequestedAmount= some number, RequestedUnits = some number, EstimatedAmount = some number where UID = 12345;
UPDATE TUNIT set RequestedAmount= some number, RequestedUnits = some number, EstimatedAmount = some number where UID = 123468;
...
This is going to be fast as it will hit primary key index and use the active partition as well;