From Silicon To Software

 

Ever Wonder What Drives Worldwide Innovation?

semiconductor applications

By Mike Gianfagna, Sr. Director, Synopsys

I’ve spent my entire adult life working in the semiconductor and electronic design automation (EDA) industries. How did that impact my social life?  For the most part, it earned me odd looks, blank stares, and sometimes a healthy dose of suspicion – you do what? C’mon, what do you REALLY do? If you’re reading this blog, you may identify with some of these comments, especially if you’ve been at it for a while. The semiconductor industry and its close cousin, the EDA industry, have lived in the shadows for most people for a long time. While the insiders of this industry lived by Moore’s law, most other folks would call a lawyer if they heard the term. Over the past decade this has all changed. In fact, the rate of change has accelerated quite a bit lately. So, what happened? Read on to find out what drives worldwide innovation.

Mike Gianfagna

The changes I am referring to are not the result of a very successful advertising campaign by either the semiconductor or EDA industries. Many may remember the Intel Inside campaign – a rather successful effort by Intel to make their name and their processor household words. The campaign worked and it was brilliant (in my opinion). But that’s not what I’m referring to. The spotlight that currently shines on the semiconductor industry is far too bright to be achieved by any, or possibly all, advertising campaigns. Something else happened. A seismic shift in our world has occurred, and there’s no looking back.

What Happened to Bring Chips into the Spotlight?

We have all watched artificial intelligence (AI) change the world around us. We now talk to our gadgets regularly and they understand us with eerie precision. Our cars will typically prevent us from making big mistakes and the requirement to set options for most things is going away – the device will figure out option settings faster than you can. But that’s not what I’m referring to, either. But it’s related to it.

Consider the figure below. You will notice a significant change in the consumer base for advanced chips. The demographic of this group has shifted from semiconductor and OEM companies to major, mainstream system companies. This shift is what has propelled semiconductors into mainstream visibility. It has also created a new generation of chip companies who are captive inside these system companies. The implications of this change are significant. These implications also present some substantial challenges. More on that in a moment.

Semiconductor companies 2021

Changing ASIC consumers

The logos in the figure above aren’t complete. One must consider other aggressive consumers of advanced semiconductors such as Tesla. The entire automotive industry, really. The recent chip shortage highlights how things have changed. If you’ve worked in semiconductors for a while, over- and under-supply are nothing new. My standard quip on the topic is that semiconductor supply and demand are in balance for five weeks every five years. Design, manufacture, and delivery of these devices requires the coordination of an incredibly complex, multi-national supply chain. The current chip shortage has made headlines around the world. Why? Because chips are now the critical ingredient for just about every new innovation (my opinion). When a critical, worldwide enabler has a hiccup in availability, the entire planet feels it.

What has propelled semiconductors into this glorious spotlight is the aggressive use of advanced technology by the companies shown in the figure above. They have all realized how critical the technology is to their future and so they have brought the whole design and procurement process in house. And that has created a new world order. Governments now discuss semiconductor technology. They discuss EDA technology, too, and consider legislation to support funding and control of these critical technologies.

Who’s Buying Semiconductors?

The new semiconductor consumers represent a very different demographic. They are less concerned about chip pricing than they are about performance, power, and time-to-results. To understand this phenomenon, let’s look at a few examples:

Apple (new Mac processor)*1

Apple was one of the earliest companies to see the strategic benefit of custom silicon. They brought processor design in house over a dozen years ago. The initial focus was on cell phone processors, and the iPhone made history. Today, they’re building processors for their computer gear. Apple’s first SoC processor for Macs represents a 119mm2 M1 die with 16 billion transistors. That’s not a typo. 16 billion.

Google (accelerate AI workloads)*2

In 2016 Google introduced the first tensor processor unit, or TPU. The device aimed to accelerate AI workloads using its TensorFlow software. Both training and inference applications were contemplated. The current TPU (v3) has a matrix architecture and can deliver up to 99% scalability in huge, 1,024-chip configurations. It handles some of the largest AI workloads in the world and consumes 450 watts. TPU v4 is expected to be available later this year and is rumored to be built in 5nm or 3nm technology. Through this massive investment, Google is committed to global leadership in AI.

Amazon Web Services*3

AWS is focused on AI in its datacenters and at the edge. Their 7nm 64-bit Graviton2 custom processor contains 30 billion transistors and occupies about 350mm2. The device is a key part of the company’s Elastic Compute Cloud (Amazon EC2). The investments don’t stop here though. Other chips in development include:

  • Graviton3 – 5nm or 3nm
  • Trainium AI training processor – one trillion calculations/sec; 30% higher throughput, 45% lower cost than GPUs
  • Inferentia Arm®-based AI inference processor; 25% higher throughput, 30% lower cost than GPUs

This small summary of what’s going on at the new semiconductor consumers tells a take-no-prisoners story. Aggressive, no-compromise projects like the ones summarized here create a vibrant outlook for both semiconductors and EDA. And finally, my non-technical friends and family understand what I do.

What Is the New Challenge for Semiconductor Design?

Earlier, I mentioned some substantial challenges with all this. The designs being built by the new breed of semiconductor consumer are very, very complex. Right at the edge of possible. You can get a feeling for the kind of complexity I’m referring to in my prior post on hyper-convergent designs.

So, we have new teams building incredibly complex semiconductor devices for new products inside a company that may be doing all this for the first time. To succeed, questions must be asked about design methodology and IP. Make sure you have all the tools and semiconductor IP you need, and make sure it all works as advertised. Ask for references and success stories. The goal is to deploy tools and technologies that support end-to-end innovation. The figure below illustrates just a few of the key requirements; there are more. So, the new challenge in semiconductor design is the need for a methodology and IP that supports end-to-end innovation.

soc design flow

Design Methodology Must Support End-to-End Innovation

Synopsys prides itself with a holistic approach to system innovation that includes all the above and more. But you probably already knew that. So now you know that semiconductors drive worldwide innovation.

In Case You Missed It

Catch up on Mike’s previous post, “Why Hyper-Convergent Designs Demand an All or Nothing Approach.”

 

References
*1 IC Insights, McClean Report 2021, January 2021
*2 IBS, Impact of AI on Electronics and Semiconductor Industries, April 2021
*3 IBS, Impact of AI on Electronics and Semiconductor Industries, April 2021; IC Insights, McClean Report 2021, January 2021