Computer Hardware – Memory : Static RAM Vs Dynamic RAM

2008-12-10

Random Access Memory plays a crucial role inside a computer. RAM is where the processor keeps the data it needs to work. The computer always has to load necessary data into the RAM because the speed of hard disks is very low for processor to work with.

Actually even the RAM is not fast enough for the processor so computers use many types of memories in different places such as cache or registers for different purposes. The purpose of this article is not to elaborate on how RAM works so without wasting time I will go into our topic.

Basically, there are two types of RAM which are Static RAM and Dynamic RAM. In this article we will compare these two types of RAM focusing on following aspects:

• How these two types are designed
• How these operate
• Issues with each type
• What is the most fast and what is most cost effective
• Where these memories are used in our computers

Dynamic RAM

This is the most common type of RAM in use today. In a dynamic RAM chip, a bit is stored in a single memory cell. Each memory cell includes a transistor and a capacitor. Thanks to the revolution of technology these transistors and capacitors have become extremely small so that a single chip can hold millions of memory cells.

How these cells operate

In a memory cell, the capacitor holds electrons to store a binary 1. To store a binary 0, the capacitor is emptied. The transistor serves as a switch to allow the memory controller of the chip to read or change the state of a capacitor. In other words, the memory controller which is a circuit operates through the transistor to either empty the capacitor to store a binary 0 or to fill the capacitor with electrons to store a binary 1.

Problem with dynamic RAM

The problem with capacitors is capacitors have leakages so those can’t keep electrons for a long time without getting discharged. So in a dynamic RAM either the CPU or the memory controller has to recharge all capacitors which stores binary 1 before those get discharged (it’s not necessary to recharge capacitors which represents binary 0 because those capacitors are needed to be empty). This automatic refresh has to take place thousands of times per second. This refresh process takes time and therefore slows down all memory operations carried out on dynamic RAMs.

Static RAM

Static RAM stores a bit in a completely different mechanism called flip-flop which does not need a recharge. A flip-flop is an advanced yet interesting topic so keeping that for a future discussion I will publish only the definition of what is a flip-flop which I believe is enough to understand the scope of this article. A flip-flop is an electronic circuit that has two stable states and thereby is capable of serving as one-bit memory. Static RAM is made of an array or a collection of flip-flops.

Since flip-flops do not use capacitors, static RAM does not need to be recharged. So this property makes static RAM faster than dynamic RAM.

Ok now what’s wrong with Static RAM?

To implement a flip-flop it takes 4 or 6 transistors so Static RAM takes more space compared to dynamic RAM leading to a less memory per chip ratio and static RAM takes more electronic components than dynamic RAM to store a bit which then makes static RAM much more expensive than dynamic RAM.

Finally, where in my computer can I find these memories?

Since Static RAM is fast and expensive whereas Dynamic RAM is cheaper and slower, the speed sensitive cache area in the CPU is made of Static RAM while the large RAM area is made of comparatively cheaper Dynamic RAM.

Was this post helpful to you? How can I improve? – Your comment is highly appreciated!

Cassian Menol Razeek