AI Spectrum

As global supply chains become increasingly interconnected and products more complex it also becomes increasingly important to understand how perturbations to this system will effect everything from supply availability to consumer demand. Understanding the complex interplay between various factors is vital for success in the global market and a field where AI can provide invaluable assistance.

In this episode, host Spencer Acain is joined by Richard Barnett, Chief Marketing Officer of Supplyframe to examine the ways AI is helping to find patterns and correlations in the vast web of data generated by the global supply chain.

For any marketplace, understanding buyer intent is vital. In the consumer space, many companies have invested in developing systems – both with and without AI – capable of doing just that. However, for B2B marketplaces gauging buying intent becomes far more difficult. This is exactly the challenge Supplyframe seeks to address with AI by understanding not just the engineering intent of companies buying electronic components but what impact events in the larger world will have as well.

In this episode, host Spencer Acain is joined by Richard Barnett, CMO of Supplyframe to discuss the ways their using AI to predict market in the electronics industry and what those predictions allow them to do.

Microchips are among the most, if not the most, complex devices ever created by humankind, packing billions of transistors into a package the size of a thumbnail. For these chips to function properly every one of those transistors must function perfectly and are rigorously verified so any problems can be corrected in future revisions. However, test data can only narrow down the cause of a problem to a certain degree and narrowing it down further is a key area where machine learning is coming into play.

In this episode of AI Spectrum, Spencer Acain is joined by an expert from Siemens EDA with more than 20 years experience to discuss the ways machine learning is playing an important role in the verification of these complex chips.

In this episode you will learn:

·        The use of AI/ML in the chip verification process (0:44)

·        Difficulties in identifying root cause (6:58)

·        Challenges of analyzing large chips (09:38)

·        Gathering ML training data (11:37)

·        The push for industry standardization (15:10)

As AI grows increasingly integrated with modern products, finding a way to quickly and efficiently design purpose-built AI accelerators for a wide range of applications is vitally important. Designing chips using High-Level Synthesis (HLS) not only allows for aggressive optimization of power usage and performance but the possibility of integrating AI itself into the design process for even greater gains. As the design process becomes increasingly interdisciplinary, HLS also offers a path to integrating electronics into MBSE workflows.

In this episode of AI Spectrum, Spencer Acain is once again joined by Russell Klein, program director at Siemens EDA and a member of the Catapult HLS team to discuss the benefits of HLS and the ways it will integrate with AI in the future.

In this episode you will learn:

·        HLS-designed accelerators vs. general purpose accelerators (00:30)

·        How HLS compares to manual optimization (03:18)

·        How AI improves on optimization heuristics (07:27)

·        Integrating chips into the MBSE process (11:39)

Connect with Russell Klein:

·        LinkedIn

Connect with Spencer Acain:

·        LinkedIn

Designing microchips is a daunting task which is growing increasingly challenging as new algorithms and software push the demand for efficient, specialized chips capable of running AI algorithms on everything from self-driving cars to edge IIoT sensors. To meet these demands in a timely manner, High-Level Synthesis (HLS) tools, like Siemens EDAs Catapult are proving themselves to be a vital tool in designing chips for the fast-passed world of AI technology.

In this episode of AI Spectrum, Spencer Acain is joined by Russell Klein, program director at Siemens EDA and a member of the Catapult HLS team to discuss the benefits of HLS and why it is playing a key role in developing the AI accelerators of tomorrow.

In this episode you will learn:

·        How Catapult can support AI (00:32)

·        AI accelerators vs. GPUs (02:32)

·        What is HLS? (04:25)

·        How HLS verifies algorithms instead of transistors (10:27)

·        Usage of HLS designed chips (11:32)

Connect with Russell Klein:

·        LinkedIn

Connect with Spencer Acain:

·        LinkedIn

Artificial intelligence has been a hot topic for the last few years as it starts to disrupt the status quo of countless industries but for EDA tools such as Solido, AI and ML have already become an indispensable proven technology. Solido leverages powerful machine learning abilities to provide answers that would normally require millions of simulations to acquire down to just a few thousand, offering a glimpse of where the AI industry may be going.

In this episode, Spencer Acain is joined by Amit Gupta, VP & GM of the Analog/Mixed-Signal Division at Siemens EDA and serial entrepreneur and founder of Solido Design Automation before its acquisition by Siemens EDA in 2017. Amit discusses why he and his team were such early adopters of AI/ML technology and the benefits of using it in the EDA space.

In this episode you will learn:

·        The role of AI in Solido (1:58)

·        The benefits of AI in EDA (4:00)

·        Validating multi-billion transistor chip designs using ML (8:32)

·        Why Solido was at the forefront of AI/ML adoption (15:16)

·        AI collaboration across industries (21:04)

Connect with Amit Gupta:

• LinkedIn

Connect with Spencer Acain: 

• LinkedIn

Even as AI drives a new level of interconnectedness between tools, it also offers the potential to reinvent the way complex physics-based simulations are run. When it comes to the use of physics informed neural networks, or PINNs for short, a number of challenges are still left to overcome, however while the road ahead for PINNs is a long one, they offer the potential for great reward at the end as well.

In this episode, Spencer Acain is joined once again by Dr. Justin Hodges, an AI/ML Technical Specialist and Product Manager for Simcenter. Justin discusses not only the ways AI is enabling connections between tools but also the challenges and benefits of PINNs and AI in physics going forward.

In this episode you will learn:

· How AI is driving connections between tools (00:32)

· How AI is changing physics-based simulation (4:40)

· The challenges of using PINNs (6:36)

· The benefits of PINNs (8:50)

· Where AI is going in the future (12:08)

Connect with Justin Hodges:

Connect with Spencer Acain: 

• LinkedIn

The products being designed and manufactured today must surpass the capabilities of what came before and then deliver them with fewer resources for various reasons, from environmental regulations to increased market competition. Making that happen is a challenging task, and even with some of the best tools, it can be difficult when relying only on the abilities of a few engineers and designers. As a result, computational resources in a digital business are becoming the differentiator for many companies looking to capture their market.

To discuss this shift and what it means for the companies adopting these new techniques, one of our guest hosts – Nicholas Finberg, a writer for the Thought Leadership team at Siemens Digital Industries Software – sat down with one of the NX product managers – Tod Parrella. Together they talk through the concept of generative design, why it’s different from topology optimization, and how it can be applied to the other challenges businesses are trying to solve.

If you are interested in learning more about the how of this process and what Siemens is doing with AI and machine learning to improve the capabilities, check out the sister podcast on the AI Spectrum series from Siemens Software – hosted by Spencer Acain.

Connect with Tod Parrella:

• LinkedIn

Connect with Nick Finberg: 

• LinkedIn

AI is not only empowering tools to function with greater efficiency and usability, but it’s also helping spearhead the next generation of interconnected technologies which will help drive further innovation through a more holistic design approach. This in turn will help parallelize the traditionally serial design process, enabling a faster design cycle and exploration of a broader design space.

In this episode, Spencer Acain is once again joined by Dr. Justin Hodges, an AI/ML Technical Specialist and Product Manager for Simcenter. Justin highlights some of the ways AI is helping build connections between different tools, and where that will lead in the future.

In this episode you will learn:

- How AI enables interconnected technology (2:18)

- How AI is evolving through cross pollination between fields (5:57)

- The ways AI facilitates the transfer of simulations and data between tools (10:47)

- How AI will help parallelize the design process (14:33)

- Knowledge capture through AI (16:55)

Connect with Justin Hodges:

Connect with Spencer Acain: 

• LinkedIn

As AI gets smarter, it will play an increasing role in designing new products. This change will take many forms, with Generative Design and Engineering being key new technologies to enable faster development in a broader design space. This, in turn, will redefine the role engineers have in the product design process from one of intensive manual work to that of orchestrators.

In this episode, Spencer Acain interviews Tod Parrella, Senior Product Manager for NX Design solutions. Tod explains the benefits of Generative Design and the challenges it faces in building trust as a new technology.

In this episode, you will learn: