23.6.2. Constructing Micro-Program Components

So, whenever we want to optimize based on encoding or compressing of the signals in each of these cells or each of the words in the program control memory, we call it as a vertical micro-program. In horizontal it is very flat and no optimization is there.

There are two primary types of micro-programs: vertical and horizontal. Vertical micro-programs focus on optimizing memory by encoding and compressing signals, which allows for a more efficient representation of control information. On the other hand, horizontal micro-programs are simpler and do not utilize any form of optimization, making them flat and straightforward. The choice between these two depends on the needs for optimization versus simplicity.

Think of vertical micro-programs like a compressed file on your computer that takes up less space and is optimized for performance. In contrast, horizontal micro-programs are like a raw image file that has not been edited or reduced in size, making it easy to view but heavier on storage.

So, basically first the most formal or the preliminary way of doing it is basically encode the signals in the control memory.

The first step in constructing micro-program components involves encoding signals into the control memory. This means organizing the data in such a way that it can efficiently represent the necessary control signals needed for program execution. For example, if a control memory word has eight outputs, we might use a decoder to manage which output is activated based on the input binary code.

Encoding control signals can be likened to using a shorthand while taking notes. Instead of writing out full sentences, you develop a system of symbols and abbreviations that represent longer phrases, making note-taking faster and more efficient.

So, I just put a 3:8 decoder then, what will be the memory size? The memory size will be 3 bits.

Using a decoder, such as a 3:8 decoder, means that you can convert 3 input bits into 8 unique output signals. This setup allows for efficient mapping of smaller input sizes to larger outputs. This compression significantly reduces the necessary control memory size, from potentially 2^n to merely n.

Imagine you have a small keychain with just three keys, but each key can unlock three different doors. By having fewer keys, you manage to open multiple doors, much like using fewer bits to control more outputs.

Now, the things will start becoming slower... the program counter out and memory address register in.

When using a vertical micro-program structure, only one output can be activated at any time due to the sequential nature of the signal processing. This means that when you want to activate multiple outputs, such as the program counter and memory address register, you must do so in separate steps. This increase in the number of time steps required can lead to slower overall processing when compared to a horizontal approach where multiple signals can be activated simultaneously.

It's like making dinner when you can only use one pot at a time. First, you boil pasta, then you use the same pot to sauté vegetables. While this is efficient in terms of using a specific pot, it takes longer than if you had several pots and could cook everything simultaneously.

So, these are clustering... which signals are simultaneously required to be 1.

The hybrid approach allows for a more flexible structure where specific signals needed at the same time can be grouped into clusters. Each cluster can have its own decoder, allowing for some parallelism while still optimizing memory usage. This method provides a balance between the vertical and horizontal approaches, optimizing processing speed while also maintaining a smaller memory footprint.

Think of this approach like organizing files in folders on your computer. Instead of having all files in a single folder (vertical), you create subfolders for related files (clusters) that allow you to access what you need faster. You still optimize your organization while ensuring similar items can be accessed together.

Consolidating what we have discussed, in horizontal micro-program less number of micro instructions are required... leading it to slower speed.

In summary, horizontal micro-programming requires fewer micro-instructions and operates faster due to its parallel structure, but at the cost of higher memory consumption. Vertical micro-programming, while more memory-efficient due to encoding and compression, tends to be slower due to the sequential nature of its operations and the increased number of control steps needed.

Consider a busy road where cars can move freely in multiple lanes (horizontal), leading to faster travel times. In contrast, a single-lane road (vertical) may be less crowded but can slow traffic significantly as it takes turns waiting in line.

--