Explore the SQL Server 2014 Buffer Pool Extension
SQL Server 2014 introduced a new feature, the Buffer Pool Extension. This functionality extends SQL Server's buffer pool, which is kept in memory with a second-level cache, backed by local SSDs of a server or virtual machine. The Buffer Pool Extension allows a larger working set of data to be kept 'in memory'. The access into the extension of the buffer pool (which is stored on local SSDs of an Azure Virtual Machine) is many times faster than accessing Azure Standard Storage. No significant advantages are expected for the Buffer Pool Extension compared to Azure Premium Storage Read Cache (which is recommended for SQL Server data files). The reason for this is that both the SQL Server Buffer Pool Extension and Premium Storage Read Cache use the local disks of the Azure compute node.
Experiences using the SQL Server Buffer Pool Extension with SAP workloads are mixed. The ideal case is that the working set required by the SAP application fits into main memory, and with Azure offering virtual machines with up to 4 TB of memory, this is usually possible. Therefore, the Buffer Pool Extension should be used only in rare cases.