>AFAIK the cache is pretty dynamic.
>
>If SQL Server thinks something needs to get cached now and memory is not available, it will attempt to free memory by dumping older cached data.
>
>I'm not 100% sure that SQL would try caching data based on a backup job, but I wouldn't be suprised.
>
>So, my guess is that the backup could indeed cause table data to become uncached.
That's what I think as well. Based on my obervations during backup time, this seems to be the case. Considering no hit was done to the table after that, during the mentioned timeframe, it could have been a factor. But, in either case, it is confirmed that after a certain period of inactivity, SQL Server drops the cache. Or, maybe it was never dropped but the reaction went the same because the backup required too much on the CPU thus SQL Server had to let go some of the cache memory.