Get SGD58.85 off your premium account! Valid till 9 August 2021. Use the Code ‘SGLEARN2021’ upon checkout. Click Here

Optimising Semiconductor Chip Placement with Machine Learning

Semiconductor chip placement is a critical step in chip design, but it is also a laborious task that can take up to weeks to complete.

FPGA (field-programmable gate array) and ASIC (application-specific integrated circuit) chips, which are widely used in consumer electronics and enterprise systems, can consist of hundreds to millions of transistors depending on their size and complexity. Before they can be manufactured, different modules have to be carefully arranged on the chip.

Given the complexity of today’s semiconductor products, this requires the expertise of highly-skilled and experienced engineers. When carrying out chip placement, the engineers work with rectangular grids, each of which contains parts of one or more modules. Poor chip placement can result in a product that does not meet performance specifications, leading to costly delays in production.

To address these challenges, chip design optimisation company Plunify decided to leverage artificial intelligence (AI) in its quest to help businesses and organisations build better semiconductor products.

Started in 2009 by two passionate engineers, Plunify seeks to improve design performance and save time and resources in the design process.

Working with AI Singapore under the 100 Experiments (100E) programme, it explored the use of machine learning to assist human experts and speed up the chip placement process. Computationally-efficient generative models were used to generate chip placements. Convolutional neural networks and other deep learning architectures were then applied to improve the quality of chip placement by predicting good placements for production.

The outcomes of the project were amazing. A ground-breaking 80 percent accuracy was achieved in applying image recognition to chip placement. This enabled Plunify’s chip design partners to slash the development cycle for products ranging from automotive power chips to 5G communications chips from 2 months to a week.

The prototyping cost for new chips was also reduced by 8 per cent, and new chip designs could be developed 10 times faster.

To find out more about our 100E Programme, please click here.

“The team from AISG provided much-needed expertise and diligence to help us design and craft the data pipeline and models for this effort. This was vital to the success of the project.”
Ng Harn Hua
Co-Founder of Plunify

Trader Turns AI engineer, thanks to the AIAP Opportunity

Whilst working as a trader in the equities and currency markets, Ngui Seng Wee was exposed to the use of artificial intelligence (AI) to make investment decisions. As he did more of that, he found his interest gravitating towards AI, algorithms and software engineering. But from a career perspective, the field of AI seemed beyond reach.

Seng Wee had graduated in 2012 with a diploma in Business Studies, specialising in Entrepreneurship and Marketing, and went on to work in a trading firm before striking out on his own, but still in the same space. “My resume would not even get me past the initial screening for an IT role,” he said.

The breakthrough
Undeterred, Seng Wee continued to pursue his interest in AI and practised applying various machine learning (ML) tools in a domain that he was familiar with – trading and investments. This helped him to understand the benefits and limitations of different ML techniques.

He also tried to pick up whatever he could about common ML tools, from books like Aurélien Géron’s “Hands-On Machine Learning with Scikit-Learn, Keras, and TensorFlow”. “It is really crucial to know what these common tools are and try to implement them in a domain that you are familiar with, so that you understand them better,” he said.

One day, Seng Wee bumped into an old acquaintance who told him about the AI Apprentice Programme (AIAP)®, and how it was willing to look beyond academic qualifications in its prospective apprentices. That was to be his breakthrough. Seng Wee applied for AIAP Batch 5, passed the technical assessment and interview, and was accepted into the AIAP.  

He believes that AIAP gave him the chance despite his lack of qualifications because they saw his willingness to learn. “If you can prove that you will make full use of the opportunity given to you, I’m sure that the AIAP team will be very happy to give you that opportunity.”

In February 2020, Seng Wee started on the AIAP, becoming the first diploma holder to join the programme.

Deep-skilling in AI
Over the course of nine months, he had the opportunity to deepen his skills in AI and gain real-world exposure to how problems could be solved.

Working on a project called Makerpose, he was involved in using a computer vision technique called pose estimation to solve real-world problems.

When he first joined the project, it was tough because the team faced challenges such as the lack of datasets or relevant literature for the use case that it was working on, and the need to balance the trade-offs between speed and accuracy in the models.

There were also many other hiccups along the way. “Things everyone assumed would work ended up not working. Problems we never knew existed started popping up.”

On hindsight, these challenges turned out to be a blessing. “I was lucky to have faced and experienced these issues while in AIAP, where I had my team and mentors to solve these problems together. Even though there was some pressure when things went wrong, it was a relatively safe environment to make these mistakes,” he said. “Because of these experiences, I became much more skilled at solving problems.”

He also picked up important skills in other areas such as communication. For example, AIAP allowed him to hone his communications skills by exposing him to people from different backgrounds. His fellow apprentices encompassed people with PhDs, start-up founders, business people, finance people, fresh graduates and some like himself who were looking for career switch.

“I learnt how to explain things in different ways to different people, so that everyone will be on the same page,” he said. “It is important that people understand what you are doing, so that the chances of a successful implementation will be much higher.”

A new career
When Seng Wee graduated from the programme in November 2020, he embarked on a new career as an AI engineer with none other than AI Singapore, thanks to the recommendation of his mentors.

He is currently with the product engineering team for Natural Language Processing (NLP), a form of AI that attempts to get a model to “understand” language for various applications such as question answering, document retrieval, sentiment analysis and chatbots.

Seng Wee likens his role as an AI engineer to that of a pilot. Just as a pilot does not need to know the in-depth details of how a plane works in order to operate and fly the aircraft, neither does an AI engineer need to have an in-depth understanding of complex mathematics. The heavy lifting is done by researchers; his role is to translate that research into solutions.

If a client comes with a problem, the AI engineer will need to know the tools available to help solve that problem. “Your client won’t care about the math; all they want to know is whether your product can help them,” he said. “AIAP equips you with the skills to find these solutions.”

TagUI : Automation Unleashed

Start your automation journey now

Last week, the dynamic TagUI development team of Ken Soh and Ruth Toh from the AI Singapore (AISG®) product engineering team welcomed new users to the Robotic Process Automation (RPA) tool via Zoom. There was interest from all over the world and it was so overwhelming that two sessions instead of one were eventually held that evening.

A lot of ground was covered in 45 minutes of presentation (you can view the slides here), which was followed by Q&A. For those who missed them or would like to refresh the material, the full video and the highlights from the second session are available below. You can jump to the parts which interest you by clicking on them.

  1. What is RPA?
  2. Demo Use Cases
  3. AI Singapore and TagUI
  4. TagUI vs UiPath
  5. Downloads Worldwide
  6. Polls Results (Dec 2020)
  7. Live Demo
  8. Concept of Reusable Assets
  9. Roadmap
  10. Navigating the GitHub Homepage
  11. Q&A

The year 2021 promises to be an exciting year as we work to promote the ease of deploying automation at work and at play. Follow us as we continue to develop the tool and grow its community of users!

The tool is very easy to use. This year we will make it 10x easier to use and we will provide support for your TagUI RPA journey for a successful deployment.

– Ken Soh, creator of TagUI

The Full Video

Video Highlights

Jump straight to the segments in the video that interest you.

1. What is RPA?

First of all, what is RPA? What can it do? See how a TagUI workflow, the set of instructions that animate your desired tasks, looks like.
( ▶️ Jump to video segment)

2. Demo Use Cases

See some demo use cases in action.
( ▶️ Jump to video segment)

3. AI Singapore and TagUI

TagUI has been with AISG since its early days. Hear why it’s a natural fit in terms of the mission of AISG and the vision of Ken.
( ▶️ Jump to video segment)

4. TagUI vs UiPath

UiPath is a leading commercial RPA tool in the market. Take a look at the value proposition of TagUI vis-à-vis UiPath and be convinced that this is the right tool for you.
( ▶️ Jump to video segment)

5. Downloads Worldwide

Who else is using TagUI outside of Singapore? A long list indeed.
( ▶️ Jump to video segment)

6. Poll Results (Dec 2020)

A recent poll sheds light on the industries and processes in which TagUI has been well utilised. The most popular features the tool offers are also revealed.
( ▶️ Jump to video segment)

7. Live Demo

Ken runs TagUI in the ‘live’ mode to demonstrate how to do a Google search query in three different ways.
( ▶️ Jump to video segment)

8. Concept of Reusable Assets

Take a look at aspects of a TagUI deployment that ensure maintainability, scalability and robustness.
( ▶️ Jump to video segment)

9. Roadmap

There is a lot of stuff to look forward to this year. Among them, the series of weekly ‘live’ Q&A sessions via Zoom has already been started. Look out for the link to the next session on this page.
( ▶️ Jump to video segment)

10. Navigating the GitHub Homepage

Ken gives a tour of the resources available on the TagUI homepage on GitHub.
( ▶️ Jump to video segment)

11. Q&A

Fielding questions from the community.
( ▶️ Jump to video segment)

Top image : Janice, a 35-year-old accounting executive, takes a well-deserved coffee break after conjuring up a robot to prepare some data required for her analysis.

mailing list sign up

Mailing List Sign Up C360