Optimality Intelligent System Explorer

As a kid, I remember reading something that passed for a joke:

We are leaving the push-button age, and entering the age where the buttons push themselves.

If you look at developments in advanced driver assistance systems (ADAS) and autonomous driving, we are not at the stage where the buttons push themselves, but we are moving in that direction. In fact, for a lot of driving, I'm not sure it will ever happen. I can imagine trucks driving on the freeway, and even on wide city streets. But near where I used to live in San Francisco, at 4th and Townsend, there is a Safeway. Its loading dock is around the back and the choreography required to stop the traffic so that a big 18-wheeler could reverse into the loading dock without hitting anything required several skilled people. My model of the future, for quite some time, is that a self-driving truck is going to require a human to do the hardest stuff, either with the pilot model based on how ships get docked or with a central operation where an operator runs many vehicles, the way drones are flown in military operations.

I think we are seeing similar leverage of human expertise in the design of electronic systems (and I mean that in a very general sense, including verification, analysis, and so on). We are moving from manual processes, to automated processes with products like Cadence Cerebrus, and towards the era in which a single designer is keeping on top of many different blocks of a design being created in parallel.

Optimality Intelligent System Explorer

Yesterday, at CadenceLIVE Silicon Valley, Anirudh Devgan, Cadence's CEO, announced the Optimality Intelligent System Explorer. The reason I didn't write about Optimality Explorer yesterday is that he also announced the Cadence OnCloud SaaS and e-Commerce Platform, to which I gave yesterday's slot on Breakfast Bytes. See my post Cadence OnCloud SaaS and e-Commerce Platform, the Next Step of Cadence's Cloud Journey.

Optimality Explorer uses artificial intelligence and machine learning techniques (AI/ML) to automatically guide optimization. The first products to take advantage of this new optimization technology are the Sigrity and Clarity solvers. Optimality Explorer takes charge of both, deciding what to change during iteration and also deciding when to conclude optimization. The designer defines a parameter to be optimized, such as return loss being less than some desired threshold. Optimality Explorer then takes design data directly from the Allegro technology, generates design variables, guides the optimization, and then automatically concludes optimization.

You can see Optimality Explorer converging on a solution in the graph above and eventually exiting once the objective value dropped below -35dB. Optimality Explorer delivers 10X (or more) faster path to an optimized design compared with traditional manual methods.

Microsoft

One pre-release user of the Optimality Explorer was Microsoft on a challenging rigid-flex PCB design. Kyle Chen, Principal Hardware Engineer discussed this:

As an early adopter of the Cadence Optimality Intelligent System Explorer, we stressed its performance on a rigid-flex PCB with multiple via structures and transmission lines. The Optimality Explorer’s AI-driven optimization allowed us to uncover novel designs and methodologies that we would not have achieved otherwise. Optimality Explorer adds intelligence to the powerful Clarity 3D Solver, letting us meet our performance target with accelerated efficiency.

As it happens, Kyle (along with Cadence's Suomin Cui) presented this at DesignCon earlier this year. You can read a deeper dive into this in tomorrow's Breakfast Bytes post.

Sign up for Sunday Brunch, the weekly Breakfast Bytes email.

.