Computer Science, Electrical Engineering
Two students completing 30 credits (20 weeks) each
In ASIC System-on-Chip (SoC) verification, exhaustive testing is critical to ensure design robustness and functionality before fabrication. Traditional verification relies on constrained random data, but achieving complete coverage, particularly for rare or complex scenarios, is challenging. This thesis investigates how machine learning, including symbolic regression and generative modeling, can predict patterns in fixed data to create randomized inputs that expand coverage. By leveraging machine learning to generalize these patterns, this research aims to optimize ASIC verification, making it faster and more comprehensive.
At Axis, we design and verify our own image processing chips that are used in most of our products. Our pre-silicon verification is done bottom-up starting from the lowest level functional unit of the chip, the module level, referred to as the design under test (DUT) in the context of verification.
The DUT's functional verification is mostly done using simulation and coverage-driven constrained random verification methodology. It simulates the DUT under random input data which is constrained by the legal configurations of the DUT, as well as steered to be able to weed out corner cases and hidden bugs. The ultimate goal is to cover all relevant functionalities of the DUT.
Coverage of particular scenarios or functionalities might require input data with patterns that are not discernible or possible to deduce. This makes it next to impossible to steer randomization with constraints to be able to cover these. To address this, Axis' verification team has developed a new methodology by using mathematical methods to generate specific fixed input data that target these cases. However, it is generally preferred to use constraint random input data instead of a fixed stimulus since randomized input data could exercise not only a specific case but also some not thought of cases.
An improvement in our methodology is to use the generated fixed input data from the mathematical models and use machine learning and other regression analysis techniques to infer a pattern which could then be expressed as a data structure or a set of constraints.
The goal of this thesis work is to investigate suitable technique(s) with the purpose of fulfilling one or more the following criteria:
The investigation of the regression analysis technique could include but is not limited to traditional regression methods, CNNs or traditional AI methods, or novel symbolic regression approaches.
We think that this thesis would be suitable for two students who have explorer spirit with strong interest in data analysis technique. Preferably, you are majoring in Computer Science or Electrical Engineering (does not mean both student are in the same major). Some experience in Python programming is also preferred.
You are valuable to us – how nice that you are interested in one of our proposals! There are a few things for you to keep in mind when applying.
For more information, contact Per Dagermo Engineering Manager at the ASIC Verification department, per.dagermo@axis.com.
Certain roles at Axis require background checks, which means applicable verifications will be done in these recruitments. Notice will be provided before we take any action.
We enable a smarter, safer world by creating innovative solutions for improving security and business performance. As a network technology company and industry leader, we offer solutions in video surveillance, access control, intercom, and audio systems, enhanced by intelligent analytics applications.
With around 4500 committed employees in over 50 countries, we collaborate with partners worldwide. Together, we thrive in our friendly, open, and collaborative culture and inspire each other to think beyond the expected. United by our commitment to inclusion, diversity, and sustainability, we consistently seek to develop our skills and way of working.
Let´s create a smarter, safer world
For more information about Axis, please visit our website www.axis.com.
Listen to Get To Know Axis – Podcast