What kind of engineers work in the semiconductor industry?

[Kun-Kun]

What does a MechE/EE/ChemE usually do in this industry?

I think that CS/CE works in the software department and architecture.

For circuit and stuff, EEs will be working on that.

ChemE works on the actual manufacturing of chips. Materials E work in this department too, I think.

Am I correct in my assessment or am I a bit wrong?

Which engineers are in high demand in this industry?
I am in Europe btw. The EU is planning on investing heavily in the chip market to become more independent from Asia. So this industry is going to reach the moon in the future.

Also, does anyone know what a “fab” job is?

Sincerely me

 

[anorlunda]

It is difficult for any engineer in any field today to not get involved with software.

 

[symbolipoint]

I was aware on one single point as a response to the topic title "What kind of engineers work in semiconductor industry". The one point I knew of was someone with Master's Degree in Physics. I never knew if he had any Engineering degree or not.

 

[Kun-Kun]

It is difficult for any engineer in any field today to not get involved with software.

Why is that? Don't EE or Computer/Software engineers get involved with software?

 

[symbolipoint]

Why is that? Don't EE or Computer/Software engineers get involved with software?

You misread that part of the post. Look: anorlunda said, "It is difficult for any engineer in any field today to not get involved with software."

He is saying that EE's and CS engineers very often get involved with software.

 

[Kun-Kun]

You misread that part of the post. Look: anorlunda said, "It is difficult for any engineer in any field today to not get involved with software."

He is saying that EE's and CS engineers very often get involved with software.

Forgive me for my mistake.

 

[phyzguy]

What does a MechE/EE/ChemE usually do in this industry?

I think that CS/CE works in the software department and architecture.

For circuit and stuff, EEs will be working on that.

ChemE works on the actual manufacturing of chips. Materials E work in this department too, I think.

Am I correct in my assessment or am I a bit wrong?

Which engineers are in high demand in this industry?
I am in Europe btw. The EU is planning on investing heavily in the chip market to become more independent from Asia. So this industry is going to reach the moon in the future.

Also, does anyone know what a “fab” job is?

Sincerely me

I worked in the semiconductor industry for a long time. I think your assessment is fairly accurate. If there is an expansion of this industry in the EU, all of the engineers you mention will be in demand. A "fab" is short for "fabrication facility", which is just the name for the factory where the chips are manufactured. ChemE's and MatE's will spend more time in the clean room, wearing "bunny suits" to prevent contamination. EE's typically are involved in design, and so don't work in the cleanroom as much. The EE design work is especially software intensive, but as others have said, all types will use software heavily.

 

[hutchphd]

But don't forget that there is a host of other skills also. Much of the fab process is photolithography at increasingly short wavelengths (because of diffraction limits) so that is all physics driven. Also there are a host of issues where surface physics is crucial. And the requirements for statistical process control must be pretty rigorous.

 

[berkeman]

A good friend of mine is a very talented ME who works for one of the larger semiconductory fab equipment vendors here in Silicon Valley. He designs the fabrication equipment for IC wafer handling and processing. Amazing stuff (I've been in one of his clean rooms -- my company sells electronic equipment to that company for the high-speed communication that goes on between the different modules in each of those IC fab machines.

 

[CrysPhys]

OP: The semiconductor industry is highly interdisciplinary. Besides engineers from various fields, it also hires, e.g., physicists, chemists, and mathematicians. There is wide overlap between field of study and responsibilities. E.g., I got my PhD in physics and worked in R&D for epitaxial crystal growth of III-V compound semiconductors. Other colleagues in the same line of work had backgrounds in chemistry, materials science and engineering, or EE.

For Si ICs (as well as other semiconductor devices), EEs also work in wafer fab, not just in circuit design. And MEs also work in IC and discrete device manufacturing (e.g., bonding, mounting, packaging, and thermal transfer), not just in design and manufacturing of production equipment.

So pick a major that you enjoy the most, and get involved with projects or internships geared towards the semiconductor industry,

 

[phinds]

Which engineers are in high demand in this industry?

Picking a career because you think a particular type of engineer is or will be in high demand is very foolish. Pick something you LOVE and everything else will follow.

 

[CrysPhys]

Picking a career because you think a particular type of engineer is or will be in high demand is very foolish. Pick something you LOVE and everything else will follow.

<<Emphasis added.>>

This reminds me of the commencement speakers who exhort new grads to "Follow your passion (heart, dream, ...), and the money will follow!" Yes, literally "and everything else will follow". But whether that "everything else" is what you had planned or not, will lead to a sustainable career or not ... depends on many variables not under your control.

OP: Phinds has a valid point. You haven't said at which stage of your education you are, and what your ultimate degree and career goal are. I also haven't followed the EU's plans that you are so ecstatic about:

I am in Europe btw. The EU is planning on investing heavily in the chip market to become more independent from Asia. So this industry is going to reach the moon in the future.

But as a cautionary tale, you should delve into the history of Sematech.

Also, be aware that the semiconductor industry, as are many industries, is highly cyclical. What is hot today as you plan your future education may be stone cold by the time you graduate and enter the work force. I've mentioned this in previous posts: I experienced (a) the semiconductor meltdown (at least in the US) of the early 1990s when major corporations and government funding agencies started shifting R&D dollars away from semiconductor devices and into software and (b) the global Internet Bubble Burst of the early 2000s (this had a massive domino effect that impacted many companies, including those in telcom, semiconductor devices, and fab equipment). As a reminder of how quickly demand for certain engineers and scientists can invert, in late 1999, there was a shortage of engineers and scientists working in optoelectronic devices [US companies were even actively recruiting overseas; and colleagues were asking me whether I had any qualified students to send their way (I was a volunteer industry mentor at the time)]. By mid 2000, there were hiring freezes; and by mid 2001, there were massive layoffs.

Get a broad background, stay flexible, and be prepared to pivot.

 

[paralleltransport]

It is well agreed upon from my coworkers is that the software industry (aka ML etc..) is currently more hot and lucrative than work in the semiconductor industry, and I work in silicon valley. I don't see any reason why this will change.

If you're after the money and flexibility as well as jobs, software will win, hands down in my opinion.

As for what do EE do in the semiconductor industry that's an incredibly broad question. There are test engineers, application engineers, design engineers, etc... I studied EE and am now working in the industry. The work is challenging technically, but also involves a lot of long hours/meeting with international partners in china/taiwan at odd times. There's a pretty large amount of grunt work especially if you work on the most advanced nodes/processes. There's a fairly large difference if you work in R&D vs. product also. Vast majority of engineers are working on product development.

Also note that the work at a fab is not as glamorous as it sounds. Many companies in the US are closing down their fabs and there are fewer and fewer (it looks like the future is dominated by very large fab like TSMC/samsung and maybe in china).

 

[anorlunda]

Don't forget manufacturing engineering and factory infrastructure engineering. If you do that for a semiconductor foundry, and then they don't need you anymore, just do it for another industry.

 

[CrysPhys]

It is well agreed upon from my coworkers is that the software industry (aka ML etc..) is currently more hot and lucrative than work in the semiconductor industry, and I work in silicon valley. I don't see any reason why this will change.

If you're after the money and flexibility as well as jobs, software will win, hands down in my opinion.

As for what do EE do in the semiconductor industry that's an incredibly broad question. There are test engineers, application engineers, design engineers, etc... I studied EE and am now working in the industry. The work is challenging technically, but also involves a lot of long hours/meeting with international partners in china/taiwan at odd times. There's a pretty large amount of grunt work especially if you work on the most advanced nodes/processes. There's a fairly large difference if you work in R&D vs. product also. Vast majority of engineers are working on product development.

Also note that the work at a fab is not as glamorous as it sounds. Many companies in the US are closing down their fabs and there are fewer and fewer (it looks like the future is dominated by very large fab like TSMC/samsung and maybe in china).

1. The OP is in Europe, not the US. He is anticipating a major European initiative in semiconductor manufacturing:

I am in Europe btw. The EU is planning on investing heavily in the chip market to become more independent from Asia. So this industry is going to reach the moon in the future.

[But, I warned the OP to proceed with caution based on this premise.]

2. Even in the US, the situation could change. The current disruption in global supply chains in general, and the shortage of semiconductor devices in particular, are causing many companies to re-evaluate their supply-chain strategy (high-volume, single-source providers and just-in-time delivery may be the lowest cost strategy, but it's not robust). E.g., instead of depending on a few large-scale facilities overseas, smaller domestic facilities (that can more flexibly pivot their production schedules) could make a resurgence: especially if the US government forks over a good chunk of change. How long lived such a resurgence might be is anyone's guess.

3. A lot also depends on the whims of the MBAs in charge of corporate strategy. I finished my PhD in the early '80s. At the time, many major corporations (e.g.,, AT&T and IBM) believed in vertical integration: along with software and services, they had extensive R&D and manufacturing in materials, components (including semiconductor devices), hardware modules and boxes, and complete systems. In the early 90's, many corporations started to shift away from this model: they could make more profits from software and services; let someone else take on the high expense of R&D and manufacturing of hardware. But then in the early 2000's, some service providers started to come to this stark realization: if they use the same hardware as their competitors, and if their competitors' software guys are just as clever as their software guys, they really can't differentiate their services other than on price: a race to the bottom line. Lightbulb moment: Hey, maybe they should develop their own, superior, hardware that will provide superior services. And that strategy holds until the next iteration of CEOs.

 

[paralleltransport]

1. The OP is in Europe, not the US. He is anticipating a major European initiative in semiconductor manufacturing:
[But, I warned the OP to proceed with caution based on this premise.]

2. Even in the US, the situation could change. The current disruption in global supply chains in general, and the shortage of semiconductor devices in particular, are causing many companies to re-evaluate their supply-chain strategy (high-volume, single-source providers and just-in-time delivery may be the lowest cost strategy, but it's not robust). E.g., instead of depending on a few large-scale facilities overseas, smaller domestic facilities (that can more flexibly pivot their production schedules) could make a resurgence: especially if the US government forks over a good chunk of change. How long lived such a resurgence might be is anyone's guess.

3. A lot also depends on the whims of the MBAs in charge of corporate strategy. I finished my PhD in the early '80s. At the time, many major corporations (e.g.,, AT&T and IBM) believed in vertical integration: along with software and services, they had extensive R&D and manufacturing in materials, components (including semiconductor devices), hardware modules and boxes, and complete systems. In the early 90's, many corporations started to shift away from this model: they could make more profits from software and services; let someone else take on the high expense of R&D and manufacturing of hardware. But then in the early 2000's, some service providers started to come to this stark realization: if they use the same hardware as their competitors, and if their competitors' software guys are just as clever as their software guys, they really can't differentiate their services other than on price: a race to the bottom line. Lightbulb moment: Hey, maybe they should develop their own, superior, hardware that will provide superior services. And that strategy holds until the next iteration of CEOs.

I think the key to recognize is that once you do EE, you are pinning yourself to one particular type of work (let's say you are an ASIC digital designer, then that's it that's your specialty).

As a software engineer (or better yet ML engineer), you can certainly work for semiconductor companies to help their infrastructure or work in the pure software industry. That's why it's a bit more flexible and resilient to changes in the industry. If the semiconductor industry improves, so will job availability and hiring for software engineers. The converse is not true for semiconductor engineers.

For example, one often thinks of Nvidia has a hardware company right? I'll bet you they are hiring more software engineers than pure hardware engineers per unit time at this moment (i consider chip validation as a software job).

 

[gwnorth]

^ I disagree. EE is a broad enough degree that you should be able to pivot from one industry to another just like software engineers. Would it require some additional training? Yes but the same is also true in software. The constraints come when you choose a more narrow and specialized undergrad like computer, or biomedical, or aerospace, or mechatronic engineering. A generalized electrical engineering degree should keep more doors open.

 

[paralleltransport]

^ I disagree. EE is a broad enough degree that you should be able to pivot from one industry to another just like software engineers. Would it require some additional training? Yes but the same is also true in software. The constraints come when you choose a more narrow and specialized undergrad like computer, or biomedical, or aerospace, or mechatronic engineering. A generalized electrical engineering degree should keep more doors open.

The original poster was interested in the semiconductor industry specifically. The view so far I've seen there pretty bleak. A generalized EE degree usually is versatile, for the same reason software engineering is versatile: if you had enough coding in your EE degree, you may be able to switch to ML or software engineering and ride the bandwagon. Most EE I graduated with don't work in the semiconductor industry anymore simply due to lack of opportunities. Why work for a hardware company when you can work for facebook/amazon/google etc.. with way more perks and salary/bonus.

The last few years, semiconductor companies have been consolidating. This is basically due to the industry maturing (this should concern you if you're an engineer). Companies can't grow anymore due to semiconductors being commoditized, and the only way to grow is to merge with the competition. It means R&D/novel product teams being laid off or put to do mundane maintenance work. Very good opportunities for the business types (M&A, reorgs etc...), less so for EE's. I'd love to be more optimistic, being a EE myself, but this seems to be the reality for several years.

 

[phyzguy]

The original poster was interested in the semiconductor industry specifically. The view so far I've seen there pretty bleak. A generalized EE degree usually is versatile, for the same reason software engineering is versatile: if you had enough coding in your EE degree, you may be able to switch to ML or software engineering and ride the bandwagon. Most EE I graduated with don't work in the semiconductor industry anymore simply due to lack of opportunities. Why work for a hardware company when you can work for facebook/amazon/google etc.. with way more perks and salary/bonus.

The last few years, semiconductor companies have been consolidating. This is basically due to the industry maturing (this should concern you if you're an engineer). Companies can't grow anymore due to semiconductors being commoditized, and the only way to grow is to merge with the competition. It means R&D/novel product teams being laid off or put to do mundane maintenance work. Very good opportunities for the business types (M&A, reorgs etc...), less so for EE's. I'd love to be more optimistic, being a EE myself, but this seems to be the reality for several years.

I have to disagree. Surely you have heard about the global shortage of semiconductors. Every semiconductor company I know is building new fabs as fast as they can. I think there will be significant opportunities going forward.

 

[Joshy]

That field looks like it's expanding to me because more companies are starting to make their own chips. They can't create discriminators if their competitors can simply buy the same chips and massage them into their product.

I feel like it's easier to switch from EE to software compared to the other way around, and it's not just because of the pay. I wouldn't suggest that anyone chooses EE for this reason, but I think there is some resilience and versatility in being an EE if you enjoy it and want to pursue a career/field it covers. A person stuck in a role is pinning themselves more than the role or career.

 

[paralleltransport]

I have to disagree. Surely you have heard about the global shortage of semiconductors. Every semiconductor company I know is building new fabs as fast as they can. I think there will be significant opportunities going forward.

Yes I've heard of the global supply chain problem. This is mostly a supply chain problem but has not a whole lot of impact on the design work. I expect there will be a short term boost of funding for hardware startups, but investors will quickly realize that doing hardware is... hard. It will return to the original level. The supply problem a lot of it has to do with old chips (chip that were designed 10-20 years ago). The design work for those chips have already been completed. It doesn't mean there's a shortage of EE designers. I assume some fabs will be built, but I doubt there's a lot of people who's dream job is to be a fab technician. The more interesting jobs are in design & R&D, and new product introduction.

The best EE jobs I know of are actually working for software companies (google, facebook, amazon etc...). You can just see all the best hardware engineers leave their old jobs at traditional elite hardware place (qualcomm etc..) to infer what is happening to the industry. But if you're going to work for a software company anyway, it's in your interest to just do what they value (software). Management will treat you much better if you are the core competency of the company. I have friends who work for FAANG as hardware engineers. They consistently mention how software is still king. Yes you may find a job there but doesn't mean they will treat you as well as their core team.

To give another perspective, if you go to any elite school job fairs, the ratio of software to hardware jobs being offered is about 10 to 1 if not even more skewed.

My prediction is that traditional hardware companies will just stabilize. The name of the game will be cutting costs to make some margins. The software companies which have all the money will build their own hardware divisions since they want to own the whole product top down and squeeze out the semiconductor companies entirely. They will absorb all the talent since they pay better and offer more perks. You'll be left working with the remaining people who either are: not good enough to switch, or just too old to change company.

The day when Qualcomm/Xilinx etc... start giving me free massage, breakfast lunch dinner, gym/pool and pay me 200k+ entry level comp package, I'll change my mind.

 

[Joshy]

I would not compare Qualcomm (pay and benefits) to other hardware companies lol. Isn’t this thread going off topic seems like a hardware versus software discussion.