editor's blog
Subscribe Now

AI for EDA: Rohit Sharma’s view from EDPS

 

Rohit Sharma is the founder and CEO of Pairpath, an EDA company with a clear mission:

“Enable customers squeeze every pico-second of performance and every milli-watt of power by efficiently providing sign-off accurate models.”

At the recent Electronic Process Design Symposium (EDPS) held at SEMI’s HQ in Milpitas, Sharma discussed the role(s) that AI might play in EDA. He started by noting that AI/ML research now consumes more than 1% of the world’s R&D budget. (Other EDPS speakers noted that the number of AI research papers has been growing exponentially, supplanting Moore’s Law for semiconductors with some other AI-research related law.)

Sharma said that the most likely use for AI in EDA was to add new features. In other words, he expects that the addition of AI to EDA will not be disruptive, but it definitely has a place. The most likely best fit for AI is in replacing algorithms that have not been successful, or not sufficiently successful.

The example Sharma gave was for cell classification—for example, characterizing a certain transistor layout as a full adder. Sharma said this is a common EDA problem and it’s an NP-complete problem. Although “NP” stands for “nondeterministic polynomial” and NP-complete are the hardest NP problems to solve, in my own mind I read “NP complete” as “not possible to complete.” At least not in any commercially practical amount of time.

It’s sort of like the dilemma that the newly reconstituted Spock faces in “Star Trek IV: The Voyage Home” (aka “Star Trek saves the whales.”) Here’s a dialog fragment from the movie to remind you:

 

Kirk: Mr. Spock, have you accounted for the variable mass of whales and water in your time re-entry program?

Spock: Mr. Scott cannot give me exact figures, Admiral, so… I will make a guess.

Kirk: A guess? You, Spock? That’s extraordinary.

 

NP-complete problems are like that. They have “high dimensionality” (Sharma’s words), so they’re hard to encode into a deterministic algorithm. AI inference used for pattern matching has no dilemma here. AI inferencing engines will happily serve up their best “guess.”

Sharma also listed the challenges associated with adding AI to EDA (generalizable to any AI use):

  1. A clear value proposition. (Just because you can use AI doesn’t mean that doing so is a good idea.)
  2. The AI use model for any specific application
  3. Data engineering. Be sure to look at the data set(s) before trying to apply ML.
  4. High dimensionality. (The Spock dilemma.)
  5. ML technology selection
  6. Integration of AI into legacy systems
  7. Acceptance of probabilistic results (will AI’s best “guess” suffice?)

In his concluding remarks, Sharma said that despite these challenges, he expects AI/ML will very likely alter the way EDA software is written.

 

One thought on “AI for EDA: Rohit Sharma’s view from EDPS”

  1. Let’s see now: EDA was born when “VERILOG CAN BE SIMULATED!!!!” became the driving force Verilog should(MUST) be used for design entry. Designers were reluctant, but as usual hype and buzzwords prevailed.

    No wonder “The example Sharma gave was for cell classification—for example, characterizing a certain transistor layout as a full adder. ” is a problem. They do not yet realize that a transistor layout is a Boolean thing, therefore it is a total mystery.

    There were countless full adders in technologies ranging from pulse gate, to nands, nors designed, built, and used before Design Automation. In fact, the origin of Design Automation (DA) was to wire PCBs for the IBM System360 in the early 1960’s.

    Automated Logic Diagrams were used for logic gates to show fan-in and fan-out for each gate.
    Starting at any gate it was possible to find the input logic conditions, the gate logic function, and the gates in the network that use the output.

    There were no simulators. so waveforms had to be hand drawn if they were needed.

    So now EDA still has trouble characterizing an adder? Carry save and carry lookahead adders were invented, eye-ball verified, pencil and paper simulated, build, used over 50 years ago.
    The key was Boolean Algebra — which EDA has nothing to do with, thankyou very much.

    I was one of the first users of ALDS. I designed, debugged, trouble shot, retrofitted from the smallest to the biggest. What? Without Verilog, VHDL, simulation, synthesis?

Leave a Reply

featured blogs
Sep 21, 2021
Placing component leads accurately as per the datasheet is an important task while creating a package footprint symbol. As the pin pitch goes down, the size and location of the component lead play a... [[ Click on the title to access the full blog on the Cadence Community si...
Sep 21, 2021
Learn how our high-performance FPGA prototyping tools enable RTL debug for chip validation teams, eliminating simulation/emulation during hardware debugging. The post High Debug Productivity Is the FPGA Prototyping Game Changer: Part 1 appeared first on From Silicon To Softw...
Sep 18, 2021
Projects with a steampunk look-and-feel incorporate retro-futuristic technology and aesthetics inspired by 19th-century industrial steam-powered machinery....
Aug 5, 2021
Megh Computing's Video Analytics Solution (VAS) portfolio implements a flexible and scalable video analytics pipeline consisting of the following elements: Video Ingestion Video Transformation Object Detection and Inference Video Analytics Visualization   Because Megh's ...

featured video

Gesture Detection for Automotive In-Cabin Applications

Sponsored by Texas Instruments

See how using 60GHz radar for automotive in-cabin gesture is ideal due to its small size and ability to sense through various materials. Applications using gesture control include changing radio stations, answering phone calls, opening windows, and more.

Click to learn more about gesture detection using 60GHz mmWave radar sensors

featured paper

Configure the charge and discharge current separately in a reversible buck/boost regulator

Sponsored by Maxim Integrated (now part of Analog Devices)

The design of a front-end converter can be made less complicated when minimal extra current overhead is required for charging the supercapacitor. This application note explains how to configure the reversible buck/boost converter to achieve a lighter impact on the system during the charging phase. Setting the charge current requirement to the minimum amount keeps the discharge current availability intact.

Click to read more

featured chalk talk

The Gateway to Connected Intelligent Vehicles

Sponsored by Mouser Electronics and NXP Semiconductors

Connectivity is going to play a vital role in the future of connected and autonomous vehicles. One of the keys to the success of our future automotive designs will be the incorporation of service-oriented gateways. In this episode of Chalk Talk, Amelia Dalton chats with Brian Carlson from NXP about the role that service-oriented gateways will play in the future of connected and autonomous vehicles and the details of NXP’s new S32G2 vehicle network processors that are going to make all of this possible.

Click here for more information about the NXP Semiconductors S32G2 Vehicle Network Processor