Thursday, May 28, 2020

What Today’s Engineers Need to Know

What does today’s engineer need to know? And how should the media such as EE Times and EDN — cover such topics for their audience?

These are overarching questions that might seem too general. Yet, such musings often expose — in the open and in public — preoccupations in the minds of the industry’s thinking heads.

Daniel Cooley, senior vice president and chief strategy officer at Silicon Labs, did not disappoint us during a recent chat. He went straight to four big industry-wide topics that he believes are changing the face of the tech world:

  1. AI
  2. security
  3. the roles that tech companies play in the real world (will be scrutinized by governments and consumers)
  4. technology stack (what happens in the cloud matters to chip designers)
4 Things Today’s Engineer Must Know

Cooley deems these topics either too poorly framed or not sufficiently explained in the media. Simplistic views on AI are already creating huge misconceptions in the market. Although engineers loathe politics, they must appreciate that the tech companies they work for are facing much more scrutiny from government regulators, just like utility companies. Lobbying in Washington might have meant nothing to most chip engineers ten years ago. Today, US chip companies can’t even sell to Huawei.

Following is a portion of our conversation with Cooley, during which he explained why these four issues are high on his mind and why he thinks engineers should be paying attention.

On AI

Daniel Cooley: First on the technology front. This is a big one. But there are too many misconceptions out there, because of the way we talk about machine learning and artificial intelligence.

[AI is covered in such a way that] it is everything to everybody, and it means nothing. I’m reading it out there and there’s a problem.

I mean, there are actually really good applications for machine learning and AI, [and they are not covered.]

AI is fundamentally a new kind of computing. It’s not ‘this if then else’ kind of stuff — that we’ve been working with for the last 40 years. This is really just fundamentally different.

And I think someone needs to sit down and parse out why it’s different and explain where it is good for that.

We know today the difference between general artificial intelligence, which is Terminator and robots, where most of this [coverage] is going right now, and [something analogous to] drug sniffing dogs.

You know, dogs get trained really well to do something. But you have no idea how it works. I can’t ask my dog, how his brain is working. [But] I don’t have to program his brain anymore. I just kind of train him.

Like understanding the dog’s limitations, not enough people are understanding what the technology is, how it’s being applied, what it’s being used for, what’s it not being used for.

They just say it’s AI. VCs are doing all this stuff, companies are starting, and every company on the planet claims to be an AI expert now even though they’re not.

So, AI is a fundamental technology. What is it? What isn’t it? Where did it start? Where did it come from? Trace it back to Carver Mead in the 1980s with neuromorphic computing. Draw the lineage.

Daniel Cooley

On security

Cooley: The second thing is about security technology. Security technology has been primarily a software and it’s been a higher layer software problem for a very long time.

But it’s now working its way down deep into the technology stack to the foundries. Actually, things are happening at the foundries, chip design houses, and [at] every level of the technology stack.

When it comes to security, I think engineers need to know why it matters.

They need to understand the implications for when it goes wrong and how it’s implemented. How is it being applied? And when it goes wrong, here’s what happens.

I think everybody understands credit card theft and the big stuff that we read about in the news because a hundred million people’s accounts got stolen.

But how does security impact chip design? How does it impact embedded software and, and what is the industry collectively doing about that?

There’s a lot of people out there talking about security, but it’s another one of those very, very noisy topics.

If you can help distill it down to something that’s manageable and meaningful, I think that’s important.

On roles tech companies play

Cooley: This isn’t so much about core technologies, but I think people need to understand a little bit more. It’s about the role that technology companies play in the world.

And this has been changing over time. Twenty years ago, the best tech companies, all they had to do was [to] be good at tech and have the best mousetrap. For example, the best search is what made Google Google. And the best cell phone technology made Qualcomm.

These big, big companies just were the best at what they did. But everything has changed in the way technology has to be connected to governments and to the trust that individuals put in these tech companies. You’re trusting Amazon with a lot when you talk to Alexa.

Technology companies will be scrutinized the way energy companies have been for a very long time. The way other industries such as pharmaceuticals, transportation, manufacturing and even entertainment have been scrutinized, right? [For example], there is a whole division of our government looking after entertainment [so that] we have rating systems on movies and TV. [And if you don’t comply], you know, you’re liable. You can’t just publish anything you want.

So, tech companies are going to have to contend with all this. You saw [what] Microsoft in the ’90s went through. That was just the first of, I think, 50 years of technology [history], fusing its way into every ecosystem.

And I have no idea how you actually take that topic and make sense of it, except that all the engineers out there [need to know what’s] happening in the background.

And whether they like it or not, [there’s an] expectation put on technology companies — from diversity to corporate social governance to investor expectations to how we lobby, through the semiconductor industry association, for example.

We might not have thought we needed the big lobbying thing 10 years ago. But all of a sudden, now we do. We’re in the news because we can’t sell chips to Huawei. And what does that mean? So, you know, general awareness needs to be there about the roles that technology companies play, and how they interact with the world. I know, Junko, you publish on that a lot and I read all your articles there.

On technology stack

Cooley: The fourth thing here that I think is needed is the technology stack — from hardware to cloud software. It is changing a lot.

The old stratification that we’ve thrived on for the last 20 years isn’t going to be the same technology stack in 20 years.

Take, for example, fabless chip design houses. We sell chips to pure-play software companies like Microsoft or Google. But it’s changing. All those companies are doing their own chips now. They’re going down-stack. All the chip companies are moving up-stack like Nvidia or Intel.

So, what becomes important is drawing some transparency, drawing some concrete examples of this… like the role that foundries play. When Globalfoundries stopped seven nanometers, that was a catalyst for a whole lot of changes. That’s still trickling through the system right now.

When we make products at Silicon labs, we have engineers at our company designing the chips, of course. We’re a chip design company and we sell chips and that’s how we make money.

But we are also doing the embedded software on those chips, writing, building products for our supply chain so that we can make those chips securely.

We have to deploy equipment into our supply chain so that we can inject information.

We have engineers writing software for the mobile phones that have to talk to these devices. We have engineers writing software for the cloud to manage the data and software updates.

So, the product itself exists across this entire spectrum now in a way that it just didn’t happen even five years ago.

That trend is going to continue. If you can cover anything on the stratification of the technology stack, it will help a chip designer understand that what happens in the cloud matters to them and their product or vice versa even.

To conclude his thoughts on “AI, Security, Tech companies’ roles in the global economy and technology stack that must be fused across all the layers in the society,” Cooley noted:

Fundamentally, computing, PCs, Internet, and smartphone got us the industry we have today. TSMCs biggest priority every year, I guarantee you, is getting the next Apple socket. Thats whats been driving the biggest foundry, the biggest cell phone company and the biggest chip company thus far. But in 20 years or even 10 years, it’s not going to be that. It’s gotta be something else. So, how do we show people where things are going in the long-term?

The post What Today’s Engineers Need to Know appeared first on EE Times Asia.



from EE Times Asia https://ift.tt/3gxjGYA

No comments:

Post a Comment

Please do not enter any spam link in the comment box.

How I channel my inner Star Trek character at work

In a recent Twitter thread , I self-identified as "some days Deanna, some days Riker." Others shared their own "Star Trek Sp...