A RAM Primer
L2 cache
Caching is the art of predicting what data will be requested next and having that data already in hand, thus speeding execution. When your CPU makes a data request, the data can be found in one of four places: the L1 cache, the L2 cache, main memory, or in a physical storage system (such as a hard disk). L1 cache exists on the CPU, and is much smaller than the other three. The L2 cache (second-level cache) is a separate memory area, and is configured with SRAM. Main memory is much larger and consists of DRAM, and the physical storage system is much larger again but is also much, much slower than the other storage areas. The data search begins in the L1 cache, then moves out to the L2 cache, then to DRAM, and then to physical storage. Each level consists of progressively slower components. The function of the L2 cache is to stand between DRAM and the CPU, offering faster access than DRAM but requiring sophisticated prediction technology to make it useful. The term cache hit refers to a successful location of data in L2, not L1. The purpose of a cache system is to bring the speed of accessing memory as close as possible to the speed of the CPU itself.
Continues...
Published as Tutor in the 10/21/97 issue of PC Magazine.