Mahesh Kailasam

The future of engineering is powered by AI & Simulation. Join this free virtual Altair event about AI for engineering to learn more. https://lnkd.in/dTdjVZvk

17

1 Comment

NVIDIA and Ansys talking about leveraging ML/AI for design and automation at #DAC61. Some key insights shared by Prith Banerjee and Sanjay Choudhry 1. help sort out manufacture design at early phase 2. Basic physics as boundary foundation 3. Real time decision making in applications in Automotive and robotic 4. Purely physical driven AI has limits on local optimization instead of global: Purely data driven AI May have accuracy calibration issue. Hybrid can provide the sweet spot. 5. Domain adaptive training to create domain specific foundation model, and the further developing small model(SLM) for particular application, such as Chip Nimo, Coollitho, etc 6. LLM should not be expected to have all accuracy needed for specific domain application. 7. There have been quantum jumps for ML computation from hardware(chip+systems) Innovation such as nvlink. 8. Massive libraries and developers is a key to the success of CUDA+nvlink+GPU integrated stack; 9. Speeding up is the name of the game on having AI work with existing automation and simulation tools 10. When GenAI is applied in chip design, the new codes/new designs will be generated, RAG will improve accuracy and relavancy. 11. AI replacing chip designers? No, just like robotics not replacing surgeons, but it will change some natures of the jobs, at least with skills of utilizing AI.

365

6 Comments

Validating perception algorithms on ADAS ECUs is key to ensuring the safety and reliability of ADAS and AD functions in the vehicle. The industry must efficiently scale test coverage by implementing simulation methodologies, which moves testing from the road to the HiL Bench. Ansys and NI (now part of Emerson) are partnering to offer real-time, physically accurate high-resolution camera synthetic data for HiL validation to address testing constraints. #ansys #avxcelerate #sensors #sotif #safad

17

2 years and 6 months into my stint at Hexagon ... .. and it's come down to this moment. On June 13th, tune in to our webinar: https://lnkd.in/ebGqtSVn where we reveal to the FEA community the remarkable transformation of Hexagon’s structures portfolio. The key highlights, also shown in the teaser video: 1) The #MSCApex evolution from a modern, direct modeller and mesher to full coverage of #MSCNastran for linear, nonlinear and topology optimization workflows  along with direct access to cloud compute.  2) The first release of a brand new high-performance nonlinear solver in SOL400 that provides significantly better performance, parallelization, and usability in mainstream nonlinear workflows. 3) Both built on the backbone of significant improvements to MSC Nastran’s core competencies to make it the #FEA software of choice for linear and nonlinear analysis built on an open framework extensible to multiphysics and model-based systems. Now, for some reflections. This journey that began early 2022 has been inspirational, emotional, and professionally super rewarding. Inspirational, because we got a “founding core” team to buy into this transformation, and propose a couple of moonshot projects around a 60 year-old code base, a 40 year old code base and a 10 year old code base.  Hats off to these visionaries for seeding these projects with their innovation, design, and grit.  Thank you Bruce, Shivakumar (Shiva), Douglas (Doug), Murali, Hemant, Fisher to call out a few. More to come! Emotional, this undertaking required resilience, conflict management, and alignment without losing sight of the goal. The x-functional challenges were worse than changing the tires of a running car. It was transforming the entire platform from an existing engine to a more efficient one and it involved humans, their emotions, tensions and aspirations. Thank you Rizwan, Luca for the coaching, the sparring, and the lessons we've all learnt along in this journey. It's only keeps getting better now! Professionally rewarding, in that it’s ok to believe in a idea. It’s ok for the team to say y’all are doing a great job. But, when your customer says why can’t I use this alpha version in production, you know you’ve crossed that chasm of self-doubt. These anecdotal moments make the late nights, early mornings, and canceled vacations worth it. That's the reward. And kudos to the direct hive Mitch, Vasiliki (Vicky), Tony, Joanna, Pramod, Vishal, Stefan, Melinda, Shank, Daniel, Jean-Daniel (JD) for daring to believe. Thanks to the field and all our early adopters for working with us. We are only getting started! Now, how does FEA matter in today’s reality of AI ... well, the question we had stacked ourselves against : “Can we #buildbetter, and #solvesmarter and supercharge our users?” The answer is ... . But, we can only reveal so much, right? Till the next #moonshot. Or, DM, learn more :) Mahesh Kailasam Hugues Jeancolas Jessica Zheng Guillaume Boisot MSC Software

166

19 Comments

It was a pleasure participating in a Fireside Chat with Sanjay Choudhry from NVIDIA on the topic of "Leveraging AI/ML in Engineering Design and Simulation". We discussed how Ansys and NVIDIA are working together on this important topic. Thanks to Brandon Wang from Synopsys for sharing the details of our conversation. #ansys #nvidia #ai #ml #innovation

126

5 Comments

This might very well change the way you design! Get insights into a large design space in minutes as easily as you would using Excel to analyze numbers. #engineering #simulation #design

30

AI-powered engineering is key to successful digital transformation. With Altair, leverage #Data from any source to build strategic digital assets for continuous improvement. Register now (https://bit.ly/3KLB4cf) to join Ujwal Patnaik's keynote at engineering.com’s Simulation Reborn Series on Tuesday, June 25 at 12:00 EDT to learn how to: - Drive innovation and efficiency with computational intelligence. - Automate CAE #Simulation workflows and utilize predictive analytics without needing a data scientist. - Enable strategic decision-making across teams to drive business growth. #OnlyForward

41

2 Comments

Have you heard of “Element Biasing” in Altair HyperMesh? This is an advanced mesh control option that allows users to reduce their overall mesh density while increasing accuracy. Check out our blog post tutorial on how to use this powerful tool: https://lnkd.in/gKxAJZJ5

13

1 Comment

"Digital Transformation: Convergence of MSBE and Modeling & Simulation - Bridging the Gap" MTSI Digital Engineering CFA Lead, Geethesh Kukkala, will be speaking at the 25th Digital Engineering and MSBE Workshop at National Harbor next Tuesday, June 25th. View full agenda on the website here: https://lnkd.in/d9Gei3yw #employeeowned #esop #digitalengineering #modelingsimulation #digitaltransformation 

37

Don't miss this week our Altair Community Magnetic Fireside Chat: 1. Explore new designs with Shape and Topology optimizations 2. Optimization including mechanical constraints In this session, you will learn to optimize your designs freely with Shape and Topology optimization considering EM and mechanical constraints. Discover new ways to run optimizations without parameters and find new shapes that consider both structural and magnetic constraints. Understand and fix any error you may encounter in your daily usage. No more loss of any project or error that you don’t understand. We will allow for 10 mins of Q&A after the presentation To register - before Thursday April 18: https://lnkd.in/e55gHaT9

15

Do you want to learn how to simulate bulk and granular materials?   Join our Altair® EDEM™ experts on Monday 1st and Tuesday 2nd July for our next round of FREE instructor led online training sessions. This training is delivered online across two, 2-hour interactive sessions and will include a basic overview deploying the Discrete Element Method (DEM) to your engineering process using EDEM simulation software.   EDEM Introduction – Day 1 This training session covers an introduction to DEM, how to set up an EDEM model, and understand the EDEM Workflow. You will also learn best practice to create Factories, and the importance of calibrating your material model.   EDEM Introduction – Day 2: This training session is a continuation of the EDEM introduction part 1, where we’ll explore the post-processing features. You'll learn how to apply EDEM base Contact Models to simulate different granular materials, from free-flowing to cohesive and stress-dependent materials. You will also have a chance to practice an EDEM tutorial during the live EDEM Demo.   Register here: https://lnkd.in/d_33JYab   #altair #edem #discreteelementmethod #DEM #training #engineering #simulation #software #onlyforward

45

1 Comment

Electrostatic discharges present real possibilities for causing damage and failures to sensitive electronic components for manufacturers. Learn how IONA Tech adopted Creo as their preferred CAD system to create the innovative StatIQ Band that monitors and reduces the risk of ESD damage. Read their story:

1

🚁 Join us on June 25th to learn the insights and tools needed to navigate the complexities of VTOL control design and simulations efficiently. Spaces are limited - Register now! #VTOL #MATLAB #AerospaceEngineering #MathWorks

4

I approach most tasks related to systems engineering and SysML models computationally, and have come to appreciate CATIA MagicDraw's OpenAPI. The learning curve for OpenAPI is steep, so for many casual users writing scripts, it can be daunting. I created the following approach that leverages your favorite IDE (and gradle) to simplify script generation.

54

6 Comments

Exciting news! Ansys has teamed up with TSMC to deliver a high-performance Multiphysics solution for AI, datacenter, and HPC chips. By leveraging Ansys' cutting-edge technology, you can stay ahead of the competition. Learn more about this groundbreaking collaboration here! #Ansys #TSMC #Multiphysics #AI #Datacenter #HPC #Technology #Innovation https://lnkd.in/dW2XMDF6

22