Programming Logic And Design

[Abigael Ortyl]

Isthere any relationship between a computer science student and DLD course? I am studying for a Bachelors of Science and Engineering degree, and we are studying DLD (Digital Logic Design) in my second semester.

We've learned so many things about DLD but nothing is related to the CS or SE. It's all about gates and related to Electrical Engineer.Is there any need of DLD in Software Engineering? While the main thing which is concern for us is Programming Languages and Computers System?

The lesson that I aim for in teaching Programming Languages and Computers Systems is that a complex system can be properly designed as a series of levels (a stack) in which each level has two properties. The first is that the current level is completely built on just the level below and nothing else. The level below was sufficient in every way to define the current level.

More important, for something like programming languages, is that each level is conceptually complete and completely consistent. Complete means that in order to program you can think at any given level and don't need to consider other things - even the lower levels. Consistent means that you won't find contradictory things at any level.

Digital logic is usually taken as the lowest level in this stack - gates. From that we can build more complex things - adders. Adding a bit of physics to that we can build actual machines that act the way the math says the gates should behave. The next level up (perhaps) is machine language (actually microcode these days, but that is just another level). Above that is assembler language, then, perhaps something like C. But each level is complete and consistent. You don't need to think about gates as a C programmer. Nor do you need to know how the Java tools were constructed and whether they are expressed in Java, or C, or assembler.

The other big idea is that as you go up the stack, the ability to express new abstractions increases. Gates and machine language have no abstraction facilities. Assembler language gives abstractions of simple operations, but little else. Once you reach high level languages you get abstractions for things like functions and objects. You get to define and name complex things and program in terms of those complex things rather than gates.

However, the big problem is that without a digital logic course you likely won't understand "state machines" and that's a REALLY big deal. And state machines pervade quite a lot of programming.

If you hear the word "protocol", there is a state machine underneath somewhere. If you have "concurrency", there is a state machine. "Parser" or "regular expression"--ayup, state machine. "Agent" or "behavior"--also probably a state machine.

Now, you can design state machines implicitly without knowing that you did. State machines are kinda common sense. But, it's often a lot easier to deal with state machines when you know what you built, can name it, communicate about it, and use tools specifically built to deal with it.

It is really useful to understand the low level operation of the machine when programming in languages like C. Understanding how the digital circuits actually do the work aids understanding of one and two's complement arithmetic, and the relative speeds of integer and floating point operations. Having a sound understanding of the fetch/decode/execute cycle is also useful if you want to develop microcode or assembler.

You might also need to program embedded systems, GPU cards, and control any number of low level devices later in your career. Gaining the understanding for this will depend on knowing the digital design basics.

Most important of all I would say that without this knowledge it is not possible to really understand how a computer works. There would always be a mystery for you regarding exactly how the computer operates.

I hope your course is teaching you skills which will be relevant if you decide to pick up whatever new technology comes along in 10 years time. Some courses might look to be aimed at people who can turn out generic code using whichever language was hot 5 years ago, and it may well be that the digital logic module is unchanged since the 1980's - but that doesn't mean it is irrelevant.

Assuming your course is generic, there are a wide range of career choices that follow the course. Developer - Full Stack might be hot in 2019, but there are plenty of other fields which need a different focus, are not electronics or standard cell design, but benefit from closer understanding of the hardware. You might be working in a cross functional team, and need to speak to hardware folks (or worse, trying to interpret a slightly incomplete specification). You might be making a trade-off between doing work on a CPU, GPU, NN, or FPGA. You might be testing CPU designs. I guess you're unlikely to end up implementing CPU architectures, but other people on your course might.

The course developer doesn't usually know in advance what field you will want to work in, and will usually be trying to avoid closing down your future options (unless the course is tightly focused on training a specific work task).

Since you're studying BSE, I'll use an example that is common to the field of enterprise software (as opposed to e.g. hardware drivers). For reference, this is a fairly common problem and by no means a fringe case.

I have two date ranges (i.e. each range has a start date and an end date). I want to know if they overlap in any way. Please write the code necessary to evaluate if these two ranges overlap.

If you want a problem situation: You're making an application for hotel reservations and you need to make sure that you don't double book a room. No two reservations can exist for the same room on the same day.

As a coder, you will work out the logic needed for each of the four cases, and then or their results to see if any of them are true. That's the basic approach to checking if one of many conditions is true.

But if you worked it out and knew how to reduce the logic (remove duplicates, negate the logic, or remove things that are invariably proven by an already existing check), you would see that it's much simpler than that:

Consider electrical engineering as the example scenario, rather than the ultimate goal of the lesson. Because you're not learning about electricity (which is obviously a major part of electrical engineering), you're only learning about logic and how to apply it.

Electrical engineers and programmers both use logical gates, and the principle of understanding them is exactly the same, even if their practical application is for different purposes.

The implementation of any realistic software system is based on electronic elements. Those elements use electricity, generate heat, take up volume, and require propagation time to operate and interconnect.

The more a CS student learns a feel for this physical electronic basis, the less likely they might be to waste time and energy (and thus money) when implementing some software solution. (CO2 emissions and all that for the "politically correct" valuation of a programming solution.)

Advanced logic design techniques using field programmable gate arrays (FPGAs), programmable logic devices, programmable array logic devices, and other forms of reconfigurable logic. Architectural descriptions and design flow will be covered as well as rapid prototyping techniques, ASIC conversions, in-system programmability, high level language design techniques, and case studies highlighting the tradeoffs involved in designing digital systems with programmable devices. This course is generally offered summers only.

Programmable Logic Devices (PLDs) are the integrated circuits. They contain an array of AND gates & another array of OR gates. There are three kinds of PLDs based on the type of array(s), which has programmable feature.

The process of entering the information into these devices is known as programming. Basically, users can program these devices or ICs electrically in order to implement the Boolean functions based on the requirement. Here, the term programming refers to hardware programming but not software programming.

Here, the inputs of OR gates are programmable. That means, we can program any number of required product terms, since all the outputs of AND gates are applied as inputs to each OR gate. Therefore, the outputs of PROM will be in the form of sum of min terms.

The given two functions are in sum of min terms form and each function is having three variables X, Y & Z. So, we require a 3 to 8 decoder and two programmable OR gates for producing these two functions. The corresponding PROM is shown in the following figure.

PAL is a programmable logic device that has Programmable AND array & fixed OR array. The advantage of PAL is that we can generate only the required product terms of Boolean function instead of generating all the min terms by using programmable AND gates. The block diagram of PAL is shown in the following figure.

Here, the inputs of AND gates are programmable. That means each AND gate has both normal and complemented inputs of variables. So, based on the requirement, we can program any of those inputs. So, we can generate only the required product terms by using these AND gates.

Here, the inputs of OR gates are not of programmable type. So, the number of inputs to each OR gate will be of fixed type. Hence, apply those required product terms to each OR gate as inputs. Therefore, the outputs of PAL will be in the form of sum of products form.

The given two functions are in sum of products form. There are two product terms present in each Boolean function. So, we require four programmable AND gates & two fixed OR gates for producing those two functions. The corresponding PAL is shown in the following figure.

3a8082e126