By Linus Low, Electronics and Computing Engineering

I teach the Very Large Scale Integrated Circuits (VLSI) module to third-year students. VLSI is offered as an option within the Micro-electronics strand.

I began experimenting with on-line discussion in 2003. My motivation for wanting to use on-line discussions as a learning tool for students was that I thought there must be a better way to engage students in the learning process than for me to lecture from prepared notes – especially for the topics that students feel are very dry and theoretical.

Instead, I decided that I would replace some of those dry lecture topics and get the students involved in creating the learning materials using Blackboard’s discussion forum. (Blackboard is our learning management System).

This new approach would be more learner-centred and as a result, the students would be more likely to take responsibility for their learning and they would have more control over the learning process.

Motivating Student Participation

What would motivate my students to participate in the on-line discussions? The first thing that came to mind was marks, of course. However, I discovered that not all students were interested in marks. Many felt that it was not worth putting in a lot of effort to earn those marks. So while marks could be used to reward the quality of participation, they do not really attract the less motivated students.

I tried other ways of attracting students’ participation such as giving them a challenging and interesting problem (I learnt recently that this is called the “Spark” for motivating engagement in e-learning).

For example, during the Gulf War, I selected articles from the Web to show how integrated circuits are used in state-of-the art weapons. That excited them – and it was relevant to the VLSI module.

I thought back to when I used to choose articles from trade magazines to show applications of integrated circuits. Unfortunately, that bored them.

Peer pressure can be an effective means of getting non-participating students to become involved. I wanted my class to divide into groups of 6-8 for online discussions. At first, I tried letting the students form their own groups. For most groups it worked out well since the students were comfortable with each other and they were able to start off immediately. There were instances when the team members would urge one another to contribute and even chastise each other for deviating from the subject matter.

However, there would always be some students who ended up together because they had not formed any group. The performance of this group was generally poor. It took a long time for them to ‘gel’, by which time the discussion would be over. For future runs, I will take a more active role in the group formation process.

On-line Discussion Learning Design

My on-line discussions are structured over 5 weeks. The learning design involves problem-solving, where the students learn the concepts while they solve a problem presented within a case-study.

• The first week uses an ice-breaker to help the students become more familiar with the on-line discussion platform. 
• In week 2, I give the students an article which outlines the main issues involved in the case study. In class, they have the opportunity to ask questions about the article to clarify any new vocabulary or new concepts. They use the online discussion forum to report what they have found out in their groups.
• The discussion then moves online during the following 2 weeks. The students’ task is to discuss a problem scenario arising from the case study.
• In the final week of the topic, the students summarise their online discussion and do a presentation. This presentation can take several different forms. For example, in April semester 2007, the students were required to do a poster presentation.

Following are some excerpts from the online discussions to illustrate the students’ experience using this learning design.

The Students’ Learning Journey

Week 1: Ice-breaker

In an online environment, the usual cues that are so important for human interaction are missing. For example, there is no eye contact, no body language to observe and no sense of being ‘together’. It is important that a sense of online community is developed early, especially if the discussion is going to span a period of several weeks.

I used an online ice-breaker so we could get to know each other better. The task was as follows:

Task 1 (Ice-breaker)

Instructor
Dear All. Welcome to the discussion board. In this first task you are to write about an inspiring moment/experience or a person who had inspired you at some point in your life. Explain briefly why you were inspired. You have one week to complete this task.

Here are 2 typical responses from students, showing their willingness to share parts of their lives.

Student 1
The persons that inspired me the most is my parent. For my dad he wakes up every morning and return home late from work. he works so hard jus to get the average salary home for the family. Although the family is not very rich, but he always nv fails to give all the family members what each and everyone wants.

And as for my mother, she nv works when i was but had to when i started growing up because of the financial difficult of my family at that point of time. Although she dun return home as late as my father, but she also needs to wake up as early as well. And normally after works she still need to return home to do house chores and cooks dinner for the whole family. i inspired them for being so responsible and had done their duty fully as a parents.

Student 2
The person who had inspired me the most in my life until now is my primary 5 form teacher. I remember that when I was in primary 5, I was in an EM3 stream class, it was the lousiest class in the whole school. My class was being look down for our mischievous behavior and academic result. My academic result wasn’t very good, sometimes even got a FAIL in some of the subjects.

But every time when I get back my exam result from Mr Ho. He will always said to the class “Even though you all fail in this exam but at least you all improve, go home and tell your parents that you all have do your best and you all have improved”. From this sentence I was inspire not to give up on my studies even though I fail my exam. If I work hard I can do it.

Week 2: Case Study Article – in Class

Before students can solve problems, they need to have some background on the issues. I have included this next task so that students will become more familiar with the concepts involved and the vocabulary & acronyms that are used in the topic.

I could have given a lecture on this topic, but I felt that a learner-centred approach would be more beneficial. Students learned the topic more effectively by finding out the definitions by themselves. Their time wasn’t wasted by going over things they already knew and they could concentrate on the new concepts. They also felt more ownership for the learning.

Following is an excerpt from Task 2 ASIC:

Instructor
What is an ASIC? Read the article on ASIC in the attached link and complete the following:

1. List terms that you do not understand in the article.
2. Discuss what these terms mean.
3. Summarise the article in no more than 100 words.

Some representative student responses:

Student 1
The terms I do not understand are ULA, HDL, standard cell, FPGA and ASSPs.:

ASSPs: An application specific standard product or ASSP is an integrated circuit that implements a specific function that appeals to a wide market.

HDL: An HDL is a standard text-based expression of the temporal behavior and/or (spatial) circuit structure of an electronic system.

ULA: A type of ASIC chip that is partially finished with rows of the transistors and resistors built in but unconnected.
Non-recurring engineering cost: The cost to setup the factory to produce a particular ASIC

Standard cell: Standard cell methodology is an example of design abstraction, whereby a low-level VLSI-layout is encapsulated into an abstract logic representation.

A chip uses for a particular application, to improve performance as ASICs are “hardwired” to do a particular job and do not incur the overhead of fetching and interpreting stored instructions. A full custom ASIC chip use a designed mask for every layer in the chip, the full custom device have total control over the size of every transistor forming every logic gate for optimum performance. ASIC performs electronic operations by maximum speed it can go and provided the circuit design is efficiently architected.

Student 2
ASIC is application-specific integrated circuit which is created for only one purpose. In a standard cell design, the manufacturer will create functional blocks and the designers will implement their designs with what the manufacturer provides. Gate array design is where the diffused layers are predefined and wafers not connected.

Full-custom design defines all the photolithographic layers and has advantages of reduced area, faster speed and disadvantages like longer time to design and manufacture. Structured or platform design is able to reduce manufacturing and design cycle time due to pre-define metal layers and pre-characterization of what is on the silicon respectively.

Student 3
ASIC stand for application-specific integrated circuit. It is an integrated circuit (IC) customised for a particular use, rather than intended for general-purpose use. Due to the advancement of technology, size of IC became smaller. Design tools improve too. Hardware description language such as VHDL are use to describe functionality of the ASICs.

ASIC are generally faster than FPGA as they can handle as complex a design. Structured or Platform ASIC cut down the manufacturing cycle time and design cycle time by pre-defined metal layers and pre-characterization of what is on silicon.

Weeks 3 & 4 – Main Discussion Topic

The students now have a chance to apply the new concepts and vocabulary (learned during Week 2) to an actual problem.

Task 3: Case Study

Instructor
MyChip Enterprises Pte is seeking your teams’ advice on how best to implement their new product into a microchip. They have these questions:

1. What design methodologies should we use?
2. Should we use the Full Custom or Standard Cells Design Flow approach to design?
3. Which design flow best supports the design methodologies in (a)?

Your task:

1. Compare and contrast the Full Custom and Standard Cells Design Flow approaches.
2. Recommend one of the approaches for MyChip Enterprises. Explain how your recommendation is consistent with the design methodologies in (a).
3. Discuss the solutions among your team members.
   Your team has two weeks to reply to MyChip Enterprises.

The students responded well to this task. One resonse was as follows:

Student 1

1. Obviously, the design methodology they should use will vary depending on their expectations of the product, or the spec of it if they are going to design it into a microchip. But for ease of process simplicity, I will recommend them to implent it using the hierarchal method, and with it, the Standard Cell approach.
   
2. The main differences between Full Custom and Standard Cells are as follows:

   COST: Effectively, the non-recurring cost for Full custom approach is much greater, since pre-defined cells are not used, and often, each individual photo lithographic layer are defined by the design team, which has to highly skilled to perform the task effectively. In contrast, Standard Cells utilizes the manufacturer’s predefined cells, making it more cost effective

   SPEED: In terms of speed, the Full Custom approach definitely outruns the Standard Cells approach, due to the designer’s efforts in fine tuning and making use of every possible square micron in the chip. The designer will also be able to integrate an analog circuit into the chip. However, in many modern day digital circuits, the usage of the Standard Cells approach is still able to meet up with minimum speed requirments, thus rendering the super speed of Full Custom futile, unless of course maximum speed is required for the product.

   COMPLEXITY: In the area of complexity, the Standard Cells approach is much simpler, since the cells have all or mostly been predefined by the manufacturer, so there is no time wasted in implementing new cells, whereas in Full Customs, where speed and area and much more critical, the complexity inevitably increases as a result of minimizing space consumption while maximizing speed.

   AREA: Full Customs approach has a smaller area, since as mentioned above, each component in placed and designed to minimize its space consumption. This in turn, also results in a decrease in the recurring engineering costs (though much lesser as compared to the non-recurring engineering costs).

   Manufacturing and Design Time: Due to the complexity of most designs done using the Full Customs approach, and its never implemented before cells and layers, the time to manufacture and the design time is definitely much higher than that of Standard Cells approach. It takes approximately 8 weeks juz to construct the IC, without the design time included.
   

3. The Standard Cells approach compliments the hierarchal methodology, since primitive leaf cells, which can be obtained from the manufacturer can be used to build the overall complexed circuits, and should any errors be found in the overall circuit, correcting the primitive leaf cells will be sufficient(much like a domino-effect). With this approach, the product can be made in the shortest possible time, with ease of error correctibilty and circuit construction.

Week 5 – Presentation

The students presented their findings from the case study via group posters.

Task 4: Posters

Your team is to create one or more posters (minimum size A0) to explain your recommendations in task 3. The posters will be displayed along the corridors of the Micro-Electronics lab so that your classmates can learn from your efforts.

Following are 3 of the posters produced by the students:

Conclusion

Asynchronous on-line discussion has been a useful tool for my students to learn more about VLSI. While completing the discussion activity, the students have developed several skills like searching for information, questioning, comprehension, analysis, collaboration, problem solving and presentation skills.

I believe that it was much more effective to take a learner-centred approach with this topic. The students were more engaged and they felt ownership for what they were learning. And the best part – it was less boring for all concerned.

Author

Linus Low Koon Teck, Electronics and Computer Engineering