Static Randon Access Memory (SRAM)
SRAM MEMORY is normally used as "cache" memory, otherwise known as "L2" cache. It also provides the work areas where your operating system (DOS, Win95, NT, etc.), and applications are loaded and perform their functions.

SRAM circuitry is a "flip-flop" circuit that allows current to move through one side to the other, depending on which ONE OF TWO transistors are activated. This circuitry is faster than that of DRAM, but it uses more power, and generates more heat.

Due to costs involved SRAM is mostly used as external "cache" memory (L2) rather than as RAM.   DRAM is the RAM of choice because it costs less and is physically smaller.

The fastest SRAM may be as much as 7 times faster than FPM DRAM, and 5 times as fast as some EDO RAM! SRAM is about TWICE as EXPENSIVE as DRAM! SRAM is physically larger than DRAM, but takes less power than DRAM.

There several types of SRAM.

  1. Async SRAM. Async SRAM has been available since about 1985 or so. It appeared as L2 Cache in 386s. It is faster than the DRAM of those days, and has speeds of 12, 15, and 20 nanoseconds. It is still faster than DRAM and EDORAM, but the new SDRAM is comparable to the fastest Asynch SRAM.

  2. Sync SRAM. This is the fastest SRAM available. But, it is inefficent when the CPU bus transfer speed exceeds 66 MHz. It is ok for most Pentium based boards, but in the future will be replaced with PB SRAM. Its speeds are 8.5 to 12 nanoseconds address/data time.

  3. PB SRAM. This type is slower than Sync SRAM on slow speed buses; however, with high speed buses (75 MHz and faster - up to 133 MHz) it outperforms the Sync SRAM. It has address/data times of 4.5 to 8 nanoseconds. This is because it doesn't have to change its delivery bursts.