Inanc,
are your 10M rows distinct ? if not, try to get the distinct records, this should reduce the number of rows you are trying to join.
do you have any null values as username? if so, please exclude those.
are you connected to your HANA environment on the same network as the PC you are running this on or are you on VPN? network latency may also contribute to the performance of this query but should be minimal.
are there any other processes running? if so, how much memory is available while you are running this query?
just a few things to consider.