Switching from Law to Integrated Circuit Design/Semiconductors at Low Cost?

[Intel 8008]

I have a bachelors and masters in law, but I've always wanted to build semiconductors. Specifically, the architecture, not fabrication. The focus is integrated circuit design, the degree electronic engineering, the job title (presumably) design engineer or electronic technician.

Correct me if I'm wrong. I understand needing to be specific when you say you want to make semis. I also read a post here saying that IC design prerequisites start at masters. I want to get this done with the least amount of paid-for schooling.

There's a recent post where an accountant asks the same thing, and it's made clear that a Beng or Bsc is needed. Here are some whacky ideas i had before reading them. Grade them 1-5.

1. Get self-taught Sam Zeloof/Jeri Ellsworth style. Fewer resources and requires godly amounts of motivation.

2. Join the royal army as an electrical technician. No debt, but might be less relevant to IC design.

3. Convince a uni to let me do a Masters with no background in EE, where my modules are focused on IC only. Saves time & money, but unlikely to be accepted.

4. Internship/apprenticeship. Free or paid, there's no debt. But hard to find. Considering audio companies.

5. Undergrad, but foundation year skipped over as I'm a Masters graduate. Least preferred option.

The sad thing is I did an EE foundation year out of school. I didn't have the knowledge and maturity I have now, so it didn't end well. But we didn't do anything related to IC.

 

[berkeman]

I'm not familiar with the UK uni system at all, but I'll try to contribute a few thoughts and questions...

Background: I left school and did a foundation year in electronic engineering.

What does that mean? How can you leave school and complete a year of school?

It's quite easy to pass a law degree

I'm impressed, especially if you passed your Bar Exam on the first try right after graduation.

I want to work in the semiconductor industry, drawing and optimising integrated circuits.

You don't "draw" IC circuits. You design and simulate and validate them against the design rules for whatever IC geometry you are targeting. This takes significant knowledge in solid state physics and in the current design toolchain.

My goal is to get a job with the minimum amount of time spent in university

To be frank, this is a bad attitude, IMO.

I would strongly prefer an internship, and have considered an electronics technician

This sounds like your best option to me at this point, based on your post.

but realistically I'm not going to master IC design in 1 year with no background. I asked ChatGPT and Claude for help, and to be an expert in IC design without going to uni is possible

Your instinct is correct; your asking AI chatbots for their opinions is a bad idea. It will take several years of intense technical study (including the solid state physics I mentioned) to get you to a point where you can contribute on an IC design team.

 

[Intel 8008]

I'm not familiar with the UK uni system at all, but I'll try to contribute a few thoughts and questions...What does that mean? How can you leave school and complete a year of school?I'm impressed, especially if you passed your Bar Exam on the first try right after graduation.You don't "draw" IC circuits. You design and simulate and validate them against the design rules for whatever IC geometry you are targeting. This takes significant knowledge in solid state physics and in the current design toolchain.To be frank, this is a bad attitude, IMO.This sounds like your best option to me at this point, based on your post.Your instinct is correct; your asking AI chatbots for their opinions is a bad idea. It will take several years of intense technical study (including the solid state physics I mentioned) to get you to a point where you can contribute on an IC design team.

Apologies fory edit. I tend to ramble, so cut down my post to its fundamentals. I guess the best use of your time would be to ask what the prerequisites for chip design are. Are you saying that chip design does indeed require a masters?

I guess you could say I passed the bar, or the UK equivalent 1st time. But you also need 2 years of work experience, signed off by a solicitor who is a training principal, and that's where you get caught out. Guys who don't make it to top firms might end up at crappy ones paid minimum wage, or less, or not at all, or paying principals for the contracts.

Sorry for feeling bitter about the education. The British system has a problem with overeducated and overqualified adults. It's not the only country, but it's not like the US where you can get jobs on high school degrees or less. Your foundation year, undergrad and postgrad are funded 1st time. High tuition fees for foreign students are pretty much what's keeping British unis afloat, because it's slowly turned into a rentier economy where the people are the resource.

What exactly is needed for IC? Am I looking at 3 years, a year in industry and a masters?

 

[Vanadium 50]

Just so I am clear, your only EE experience was a "foundation year", in which you did poorly, and you want someone else to pay for your education doing what you did poorly. Sounds tough.

Most of the good IC designers I know have MS or higher degrees, and a lot of continuing education. You want to minimize the time in school, this is not the career for you,

 

[berkeman]

I guess the best use of your time would be to ask what the prerequisites for chip design are. Are you saying that chip design does indeed require a masters?

Yes. V50 and I are in agreement that IC designers generally have MSEE or PhD level qualifications. You need to be knowledgeable in Solid State Physics and Intro Quantum Mechanics and advanced EE subjects like circuit design and IC validation. You also need to be very skilled in logic design languages like Verilog for digital circuits, and you need to be an expert in SPICE simulations for analog circuits. It also helps if you have a lot of experience in IC packaging now, since modern IC packaging technologies are incredible and contribute a lot to the economies of IC fabrication and use.

Buckle up, buttercup!

 

[Intel 8008]

Just so I am clear, your only EE experience was a "foundation year", in which you did poorly, and you want someone else to pay for your education doing what you did poorly. Sounds tough.

Most of the good IC designers I know have MS or higher degrees, and a lot of continuing education. You want to minimize the time in school, this is not the career for you,

I have 2 arguments for that. 1: the foundation year was on broad physics. Thermodynamics and friction coefficients aren't that relevant to IC design.

2. I'm pretty sure as a masters student, I would outperform my 18 year old self, regardless of whether I took A-level physics and maths or not. Remember that the foundation year is for those who didn't study physics and maths, so we're comparing a law bachelors masters to an 18 year old still in school studying something else.

I understand there's a debate about loan forgiveness in the US, but don't drag that into this UK case. I didn't say anything about getting someone else to pay for my degree. I'll pay back the money owed, but the idea is to reduce the amount that needs to be borrowed first.

 

[Intel 8008]

Yes. V50 and I are in agreement that IC designers generally have MSEE or PhD level qualifications. You need to be knowledgeable in Solid State Physics and Intro Quantum Mechanics and advanced EE subjects like circuit design and IC validation. You also need to be very skilled in logic design languages like Verilog for digital circuits, and you need to be an expert in SPICE simulations for analog circuits. It also helps if you have a lot of experience in IC packaging now, since modern IC packaging technologies are incredible and contribute a lot to the economies of IC fabrication and use.

Buckle up, buttercup!

Our of interest, what's the rough age distribution of these guys. 4-5 years in uni and you'll start working at one of the firms at 23, assuming your career advisor knew what the hell your were talking about. I know there's probably an old chief of staff on every project at a fables IC, but would an early 30s person look out of place at an entry level position there? And before you say it, I know age doesn't matter, people start their careers at different times, and there's no shame in being a little older.

And one last thing. I know the chatbots are crap, but they give non-judgemental information and are infinitely patient, so I always find them a good starting point before jumping on forums. Plus they have the whole Internet inside their training data. You just have to tease out the right data, e.g. "let's think through this step by step to avoid making mistakes".

 

[Vanadium 50]

I get the picture. Two things to consider:

(1) Getting external support for 2nd degrees is the exception, not the rule. Getting this after poor performance is rarer still. Yes, you have some fine excuses - but what limited resources are likely to go to people with accomplishments/

(2) As I said, this will require a lot more schooling than you anticipate. And it will never stop.

I think your best option might be to discuss with your employers branching into patent law. That might provide a source of external support while you get your degree.

 

[berkeman]

Our of interest, what's the rough age distribution of these guys. 4-5 years in uni and you'll start working at one of the firms at 23

I've worked with IC designers who ran probably from 30-50+ years old. It's not just the degree, it's the experience and on-the-job learning that you get from your Mentors. We did hire a few extraordinary candidates fresh out of their Master's degree (to get young blood into the R&D organization to help us be more well balanced overall), and they were Mentored well and developed over their first few years into high-performing contributors.

And one last thing. I know the chatbots are crap, but they give non-judgemental information and are infinitely patient

We do not allow AI chatbot content in the PF technical forums, and there is a good insights article that covers some of the problems in relying on them for too much:

https://www.physicsforums.com/insights/why-chatgpt-is-not-reliable/

I think your best option might be to discuss with your employers branching into patent law. That might provide a source of external support while you get your degree.

I think that's a great idea.

https://www.physicsforums.com/threads/anyone-considering-a-career-as-a-patent-attorney.577550/

 

[Vanadium 50]

It's not just the degree, it's the experience

This.

Every so often, a Verilog expert comes up with a design and a senior engineer says "It won't wiork because of X, Y and Z." And sometimes people ignore this and plow ahead. Sure enough - it didn't*. I can think of zero cases where the engineer was wrong and Verilog was right: this is harder to do than it looks.

* In one case, the chip went into latch-up and drew so much power that it burst into flames, Usually things are not this spectacular, though.

 

[berkeman]

In one case, the chip went into latch-up and drew so much power that it burst into flames

Very cool. Wish I could have seen that one. Most of our IC first article failures were just <no workie>...

 

[Intel 8008]

I see. I don't have a problem learning on the job. That's a requirement of all jobs. I'd do it tomorrow if invited. My problem is I come from a country where 1 in 2 adults go to university. Grade inflation pushes them out of the job market and into endless internships or precarious work like uber and deliveroo. I'm a big believer in vocational skills, because the ones who came before me are proud of their 2 degrees they don't use, while the vast majority of their income comes from real estate.

0% of my law bachelors is used. The masters/bar is a course specifically designed to teach practical skills, before going on and getting 2 years of practical skills (that in my experience so far are quite different from practical skills taught). I believe the best way to learn is on the job with competitive and preferably international firms, which sums up IC pretty well.

I'm going through the patent law thread. Now I'm regretting not doing a patent law module. But then again, if most of those lawyers have 2nd degrees in science and engineering, I don't stand a chance. City of London lawyers mostly focus on insolvency, litigation, private equity, mergers and acquisitions. Basically moving money around, not science.

What would the two of you say to aspiring IC engineers to (a) encourage them and (b) ward then off? What would you like to have known? Because it seems like a job you discover over time.

This was in my original post, but I had a vague idea that I would need an EE degree at 18. However, because I couldn't point and say "IC engineer", and career advisors can only spell 'medicine' and 'law', I was encouraged to take a 'safer' option that's yet to pay dividends. Meanwhile, I would've known by my 2nd or 3rd year of an EE degree the path to take and name of the role.

Another thing, how many under 20s get into it because they're fanboys of Nvidia or some other gaming brand? And is it a bad motivation? I can't think of a reason why someone in the past 20 years would wake up and want to do IC design that doesn't involve a love of gaming PCs or consumer electronics.

Requirements so far:
IC packaging
Spice simulations
Verilog
Circuit design
IC validation
Quantum mechanics intro
Solid state physics

Theoretically, would a lean course that taught just that get you up to speed for IC design? Because if you have to wade through the elasticity of rubber to get to it, that's pretty discouraging. Not knowing about the speed of sound underwater can cut you off from the very narrow field you want to study. Not to mention the real learning that's done on the job, as mentioned.

 

[berkeman]

What would the two of you say to aspiring IC engineers to (a) encourage them and (b) ward then off? What would you like to have known? Because it seems like a job you discover over time.

I think that most of the IC design engineers that I know got into the field because in the last couple of years of their undergrad they realized that they were bright enough to pursure the harder EE specialties, and they liked the idea of being able to turn their work into ICs that sold in the millions to customers which would make good money for them and their employer.

One thing that helps a lot is that some schools (like my grad school, the University of Michigan) have small IC fabs as part of their campus lab facilities. It's hard to beat being able to take classes in IC design, and actually doing a circuit design (using SPICE and Monte Carlo simulations against the design rules, etc.), and then laying out the IC and fabricating some of them on a shared wafer with your classmates' projects and then packaging them up and testing them against your simulations. Great stuff. (One of my grad school roommates was in that program, so I got to see it all first hand, including the trials and tribulations of the design/simulation parts).

Another more recent development is the availability of microcontroller (uC) and FPGA design kits that allow you to use Verilog (or VHDL, but preferably Verilog) to design your own FPGA modules to attach to your uC development boards to do amazing functions at high speed. You can design FPGA modules to do DSP and other high-speed processing of audio/video/whatever, and use the uC interface to present the results to the end user. That is great practice in designing ASICS, which is a big step on the way to designing custom ICs.

Meanwhile, I would've known by my 2nd or 3rd year of an EE degree the path to take and name of the role.

Another thing, how many under 20s get into it because they're fanboys of Nvidia or some other gaming brand? And is it a bad motivation? I can't think of a reason why someone in the past 20 years would wake up and want to do IC design that doesn't involve a love of gaming PCs or consumer electronics.

I don't have experience with that fanboy fanbase, sorry.

 

[Vanadium 50]

Just to manage expectation - your first job will not be "design Nvidia's latest GPU" or "Design the next-Gen Intel CPU." It will be more like "We need a 555 that runs on 1.2 V".

 

[Intel 8008]

Just to manage expectation - your first job will not be "design Nvidia's latest GPU" or "Design the next-Gen Intel CPU." It will be more like "We need a 555 that runs on 1.2 V".

It's OK, my expectations are in check. My plan was to start with small UK firms, ones who do small microcontrollers for things like remotes or car parts. Then move onto large UK firms like Arm or Imagination, before fulfilling the dream of working at one of the large US firms (AMD/Apple/Intel/Nvidia/Qualcomm) or a startup like (formally) Cerebras. I wasn't going to be the next Jim Keller overnight.

I think that most of the IC design engineers that I know got into the field because in the last couple of years of their undergrad they realized that they were bright enough to pursure the harder EE specialties

It sounds like chance. Like no one starts wanting to do it. Another thing I'm keenly aware of is what the national industries are. If you're US or Taiwanese, you have an important semiconductor base. If you're German, you build cars. If you're British, you launder money. Hence the focus on mechanical engineering in Germany, and law/finance here.

One thing that helps a lot is that some schools (like my grad school, the University of Michigan) have small IC fabs as part of their campus lab facilities. It's hard to beat being able to take classes in IC design, and actually doing a circuit design

You're right, I should apply to unis with IC Labs. I can only hope that dedication and knowledge of what I want beforehand gives me a leg up over students who chance it. My worry is the safety net of 1st degree financing is gone. If I'm not good enough, I'll find out in my 2nd or 3rd year, while unis are happy to guzzle my money.

ICL presumably has the best IC masters in the UK, and they want a 1st in EE with an average of 75% https://www.imperial.ac.uk/study/courses/postgraduate-taught/analogue-digital-circuit-design/. Although, I think it's good to gatekeep degrees when the university industry takes on and fleeces 2:2 students like sub prime mortgages back in '08.

What should I prepare for beforehand? I'm thinking whether it has to be now, or if I take up a year to build up some knowledge on the side. I don't want to rush into things again. I knew someone who joined back when the fees were £3000, and treated it like a daycare with multiple repeat years. Still paid less than me overall, though.

Sounds silly asking on a forum, but I had someone 5 minutes ago conflating electronic engineering with IT, and asking whether a £500 coding boot camp would get me up to speed. There's an extreme amount of inertia pulling me away from it, since nobody IRL knows anything about engineering, let alone electronic.

If all else fails, I guess there's patent law.

 

[Vanadium 50]

My plan was to start with small UK firms, ones who do small microcontrollers for things like remotes or car parts.

Do you know who these companies are? Do they even exist?

I don't know about automotive, but I will be buying some COTS ultrasound chips. The two big names are Texas Instruments and Analog Devices, each with ~25,000 employees. You're looking for a very specific job which may not even exist.

I probably lied in my last message. Your first job will probably not be "tweak this existing chip". It is more likely to be "answer technical questions from our sales department." Oh, and Germany makes ICs - there is an enormous fab in Dresden.

 

[CrysPhys]

I think your best option might be to discuss with your employers branching into patent law. That might provide a source of external support while you get your degree.

I think that's a great idea.

https://www.physicsforums.com/threads/anyone-considering-a-career-as-a-patent-attorney.577550/

'm going through the patent law thread. Now I'm regretting not doing a patent law module. But then again, if most of those lawyers have 2nd degrees in science and engineering, I don't stand a chance. City of London lawyers mostly focus on insolvency, litigation, private equity, mergers and acquisitions. Basically moving money around, not science.

f all else fails, I guess there's patent law.

Hit the pause button on this. I'm a PhD physicist, also a US registered patent agent (in the US, working as a patent attorney requires a JD law degree, working as a patent agent doe not). I contribute to the forum thread cited above.

* That thread is dedicated to careers in patent law in the US. Qualifications and requirement for careers in patent law vary substantially in different countries. In particular, in many major countries, the qualifications and requirements (such as exams and apprenticeships) are more stringent than in the US.

* Post-BREXIT, the UK is no longer a member of the European Union (EU), but it is still a member of the European Patent Organization. If you want a good career in patent law, not only do you need to meet the qualifications and requirements to practice before the UK national patent office, you will probably also want to meet the qualifications and requirements to practice before the European Patent Office. And all of this has become messy as of June 1, 2023, when the EU Agreement on a Unified Patent Court went into effect. Most, but not all, EU members have ratified the Agreement. Post-BREXIT, UK is not an EU member and not a signatory of the Agreement. Before you contemplate a career in patent law in the UK, you need to sort all this out. Business among European countries (whether formally members of the EU or not) is closely tied; and being able to practice before a single national office might not be sufficient for a viable career. Check carefully.

* In terms of degrees, you are working backwards from all the patent attorneys I know (that includes US, Canadian, and European). I suppose if you look hard enough you will find kids who say, "When I grow up, I want to be a patent attorney." But the patent attorneys I know all started out wanting to become scientists or engineers; then, for various reasons, went to law school and became patent attorneys. E.g., I know around 6 who got their bachelor's in computer science with dreams of cooking up the Next Big Algorithm, but instead ended up writing line after line of mind-numbing code. After a year or two they headed off to law school, instead of grad school for further degrees in computer science. That would be akin to an IC design engineer with dreams of designing the Next Big Gaming Engine, but ending up working on something more mundane instead (as V50 posted above).

* Also, if you think you can satisfy your technically creative urges by working as a patent attorney instead of an engineer, think again. You are paid to draft and prosecute patent applications (or litigate them if you go into patent litigation), not to have fun working with cool inventions and learning from scientists and engineers. The billing clock is constantly ticking away. I was able to do that more than most patent attorneys or patent agents, but that's because I had a PhD Physics with 20+ yrs industrial R&D experience.

 

[Intel 8008]

Wow, you guys are really selling it. So neither IC engineers nor patent lawyers start out wanting those roles.

Let me get this straight. This IC dream job:

-might not exist
-wasn't on the radar of the people doing it when they started
-favours experience and occasionally branches out to youngsters
-requires a 1st or high 2:1 in the undergrad to apply for a masters, meaning you have to ace EE modules unrelated to IC to keep your average up
-requires specific unis who may not even have fabs to test designs
-doesn't have great pay, despite the importance of semiconductors
-isn't very UK-centric (we have a fab in Newport that focuses on GaN and SiC, but it got into trouble for Chinese investment lately)

Instead of the engineer building supercars, it's more like building a road or bridge. Glassdoor says they make £42K, and £60K+ when senior. So less than a high-end doctor, and much less than a quant with the same maths skills. I see now why people chase software engineering. Alright, I get the message. There's a reason no-one talks about or pushes IC design.

You know who I blame, besides myself? Asianometry, Fabricated Knowledge, SemiAnalysis and TechTechPotato. They make semiconductors look so easy and sexy. Really, they're just a necessity, and the engineers who stand out are 30 years ahead of new recruits. Asianometry did marketing in Silicon Valley. TTP wrote for AnandTech. The other two handle finance. What they all have in common is none are engineers. So I ask the actual engineers, and they tell me they didn't want to do it.

OK, I get the message. You don't choose IC, IC chooses you.

Patent law seems good at £82K, but it's in line with other high-end law jobs, except they don't require a Bsc. It was my dream to have transferable skills and work anywhere, so these hurdles to the EPO aren't pleasant. It's a shame you also don't mess with the technology. It seems the money follows the most commission (PE, M&A), not the most educated.

Thanks for putting these foolish ambitions to rest.

 

[Vanadium 50]

Would you rather be surprised or informed?

 

[StatGuy2000]

@Vanadium 50 , @CrysPhys , it sounds to me that both of you think that the dreams/desires of the OP (who is based in the UK) in working in semiconductors is hopeless, and that he has very little to no chance whatsoever to leave the legal field to pursue a career in engineering.

Furthermore, in the case of @CrysPhys , your implication regarding patent law is that (as necessary as patent law may be) it is a very dry and not very fulfilling field.

Am I wrong in my assessments above? If so, explain why.

 

[CrysPhys]

@Vanadium 50 , @CrysPhys , it sounds to me that both of you think that the dreams/desires of the OP (who is based in the UK) in working in semiconductors is hopeless, and that he has very little to no chance whatsoever to leave the legal field to pursue a career in engineering.

Furthermore, in the case of @CrysPhys , your implication regarding patent law is that (as necessary as patent law may be) it is a very dry and not very fulfilling field.

Am I wrong in my assessments above? If so, explain why.

I certainly didn't say anything like that. In fact I only addressed the issue of becoming a patent attorney (particularly in Europe) as a fallback position. With respect to patent law, it's not a substitute for R&D or engineering. You can't live vicariously through the creativity of others. Whether it's fulfilling or not depends on the individual, and it's relative to other options. As I discussed above, many become patent attorneys because their initial careers were unsatisfactory.

 

[StatGuy2000]

I certainly didn't say anything like that. In fact I only addressed the issue of becoming a patent attorney (particularly in Europe) as a fallback position. With respect to patent law, it's not a substitute for R&D or engineering. You can't live vicariously through the creativity of others. Whether it's fulfilling or not depends on the individual, and it's relative to other options, as I discussed above.

Fair enough. My impression was based on your observations that people who become patent attorneys (whether in the US or elsewhere in the world) tend to have ambitions of becoming scientists or engineers, and then get redirected to patient law.

I had then made the conclusion that patent law is not in itself particularly fulfilling and is thus a fallback option for those who majored in STEM and not wish to continue directly in the field. Admittedly, my assessment is in no small part based on my own biases against the legal profession (which I consider to be inherently uninteresting or unfulfilling).

 

[CrysPhys]

So neither IC engineers nor patent lawyers start out wanting those roles.

OK, I get the message. You don't choose IC, IC chooses you.

I don't understand how you come to those conclusions with respect to IC design. Berkeman wrote:

I think that most of the IC design engineers that I know got into the field because in the last couple of years of their undergrad they realized that they were bright enough to pursure the harder EE specialties, and they liked the idea of being able to turn their work into ICs that sold in the millions to customers which would make good money for them and their employer.

<<Emphasis added.>> At least in the US, choosing your future career in the last couple of years as an undergrad is not late in the game.

 

[CrysPhys]

Fair enough. My impression was based on your observations that people who become patent attorneys (whether in the US or elsewhere in the world) tend to have ambitions of becoming scientists or engineers, and then get redirected to patient law.

I had then made the conclusion that patent law is not in itself particularly fulfilling and is thus a fallback option for those who majored in STEM and not wish to continue directly in the field. Admittedly, my assessment is in no small part based on my own biases against the legal profession (which I consider to be inherently uninteresting or unfulfilling).

My perspective is that patent law is more fulfilling than many STEM careers with only a bachelor's; and has a more certain and stable path (and likely better financial return) than a master's + X yrs experience or a PhD.

In the US, there generally is no undergrad course of study in law. A "pre-law" major can be just about anything, including English lit and art history. If you're planning a career in law (in general), why pick a difficult major such as physics or engineering? Many students choose a STEM major with expectations of doing challenging, creative work after only a bachelor's. After a year or so on the job, reality sets in. Then it's either go for a grad degree in STEM or an alternative degree, such as law.

As usual, there are exceptions. One of my colleagues got his PhD in high-energy physics. His thesis work was originally planned to be experimental. But excessive delays in getting access to an accelerator caused him to switch his thesis to computer analysis of existing data. When he graduated, openings for high-energy physicists were especially rare (one major project had just been cancelled), and he found employment as a software engineer. He found the job unfulfilling, headed to law school, and became a patent attorney.

Other exceptions include senior scientists and engineers who became inventors, filed for patents, got interested in patent law, and decided to make a career change. In the good 'ol days, when Megacorps such as AT&T and IBM were flush with cash, some companies even picked up the entire tab for select scientists and engineers to get their law degrees and then work as in-house patent counsel.

But even with these exceptions, the original plan was a career in STEM.

 

[berkeman]

Wow, you guys are really selling it. So neither IC engineers nor patent lawyers start out wanting those roles.

But you can check to see if you really like some aspects of IC design...

Another more recent development is the availability of microcontroller (uC) and FPGA design kits that allow you to use Verilog (or VHDL, but preferably Verilog) to design your own FPGA modules to attach to your uC development boards to do amazing functions at high speed. You can design FPGA modules to do DSP and other high-speed processing of audio/video/whatever, and use the uC interface to present the results to the end user. That is great practice in designing ASICS, which is a big step on the way to designing custom ICs.

You can buy these development boards for very reasonable amounts of money (less than $200 USD total probably) and start designing some fun hobby systems/projects. That will get you a long way into understanding how you prototype digital ASICS, and how that leads into some aspects of IC design. And if you find yourself getting goosebumps as you code up Verilog functions in your FPGA to blaze through input data and stream the results at high speed to your uC for further processing, then you will probably know what you want to do with the next few years of your life.

(and if you find it boring and tedious and hard to deal with the details and debugging, that will also tell you something about your near future work focus)...

 

[Vanadium 50]

you think that the dreams/desires of the OP (who is based in the UK) in working in semiconductors is hopeles

Please don't put words in my mouth.

First. that's not what the OP said. He doesn't want a "job in the semiconductor industry". He wants to design ICs for small companies, growing to larger and larger companies. He said he would not be the next Jim Kellerr overnight (emphasis mine), That's a pretty narrow path.

Then he says that while he's OK with OJT, he is not so interested in formal education after he gets his new degree. Fair enough, but that's not really how that field works - and the path narrows further.

Also, he's looking at what, 4 or 5 more years of schooling? Somebody else should pay the bill. The employer, hoping to move into patent law? Nope. Narrower and narrower.

Yes, the OP is free to paint an itty-bitty target if he chooses. But if we point out that it's an itty-bitty target, we're horrible dream-crushers.