The Evolving Role of Hardware Engineers in an Era of Mature Demo Boards

Click the above “Embedded and Linux Matters”, select “Top/Star Public Account”

Benefits and valuable content delivered promptly

Is there still a future for hardware engineers?

Do you agree that the prospects for hardware work are mediocre, just enough to get by?

Netizen 1: After 12 years in hardware, I’m now a chip manufacturer FAE. Still doing miscellaneous tasks.

Netizen 2: The key is that hardware salaries are capped, low pay, and too many miscellaneous tasks.

Netizen 3: Hardware engineers rely on accumulation, but in China, those doing well tend to aim for management because it’s more lucrative, making it hard to find experienced engineers, who are mostly at manufacturers or leading system companies. Most companies are using freshly graduated young people to draft solutions, which makes it hard to improve quality.

Netizen 4: Domestic solution providers are mainly software-heavy, and it’s hard to find high-level hardware talent. The investment in equipment is large, and many in-depth tasks are better off being handled by manufacturers, so don’t expect them to solve overly complex issues.

Netizen 5: Given the salaries offered by domestic hardware manufacturers, and the required R&D speed along with the vast responsibilities expected of hardware engineers, those who can replicate public versions are considered excellent hardware engineers.

The Evolving Role of Hardware Engineers in an Era of Mature Demo Boards

A lot of people say that hardware engineers rely on experience, need to analyze requirements, design based on needs, and can handle high-speed circuits, etc.

So what if they can do it? Earn a high salary, marry a beautiful and wealthy partner, and reach the pinnacle of life? Don’t be ridiculous, just search for hardware job postings, and you’ll find that the requirements for high-speed work are extremely rare, and even this one rare position pays less than ARM.

The domestic environment is like this: technical bottlenecks, primarily based on ARM, with low-end demand being predominant and high-level demand being scarce. Compared to the internet, it can only be described as pitiful. Everyone is working with ARM, and when it comes to competition, it’s about being able to draw schematics, do wiring, and write BSP. In the past two years, there are even requirements for structural work, basically allowing one person to handle the entire product.

Hardware engineers are a peculiar kind of profession: most of those in this industry in China have never really entered it. Those who claim that the increasing maturity of upstream chip manufacturers’ DEMOs has turned hardware engineers into “paper-pasters” who copy reference designs everywhere, are indeed like that themselves, and they have never seen what a truly senior hardware engineer looks like.

The high salaries in the internet sector have attracted many talented individuals, leading to a general decline in the overall talent level in the hardware industry, which exacerbates the first point. Actually, I have a good outlook on the mechanical industry; the more talent is drawn away by the internet, the greater the demand.

You can’t ask those who are forced to work in mechanics or hardware; they are mediocre, afraid they won’t learn algorithms if they switch to software, and they will certainly tell you that hardware is not good, with deep pits everywhere, while XXX does software with a starting salary of 50K/month…

The qualification requirements for hardware experts are very high; without certain qualifications and a degree of perseverance and hard work, with some experts guiding you at the start, it’s almost impossible to achieve anything. If studying software requires a strong foundation in mathematics and logic, studying hardware also requires physics and specific industry-related engineering knowledge.

Using Kalman filtering to implement sensor data acquisition under interference conditions for process control systems, should we use signal cables with shielding drive technology, or is optical or magnetic isolation more reliable? These questions seem to go beyond just circuit signals; they are related to cost, materials, application reliability, and code compatibility.

A good hardware engineer seems to be this kind of expert: they strategize, know every technical detail, and can quickly identify potential sources of any problem, skillfully balancing cost, functionality, performance, and customer experience.

Back to the point

If you have above-average talent, doing anything is fine, not just hardware. If you have average qualifications, do something that average people can also earn money from, like software. The demand is high, and there’s always room for a few mediocre people, especially since average people can write adequately.

Actually, for “R&D engineers,” those who deserve this title are those who design new products and create new value for us. Shouldn’t they be the top 5% smartest in the crowd? Honestly evaluate yourself, and you’ll know whether you’re suited to be an engineer.

I find hardware very interesting; at certain strategic levels, hardware design always requires people of scientist-level caliber to handle it. If you have passion, you might want to give it a try.

In the end, the essence of this question is problematic. Most people answer it according to the unspoken intent, which is quite interesting. The true meaning of this question is, “I just want to go to work like everyone else, work hard, and learn; can I get excess returns?” Quick money must have a reason behind it: either you’re particularly smart, particularly hardworking, or you have exceptionally good luck.

Some people say that most demands can be easily met by copying DEMOs. I feel that “doing” is like this, but “finalizing” may not be possible. Have you not seen so many products that malfunction under interference data or collapse under high or low temperatures, or those with poor accuracy, slow response, or that occasionally freeze and need to be restarted? Where do these defective products come from?

It must be noted that the main purpose of DEMOs is to demonstrate “technical feasibility,” and its biggest problem is that it does not integrate with specific industry applications. Some industry-customized DEMOs only consider the technology itself without accounting for factors like vibration, interference, and environmental temperature and humidity.

And this is precisely the greatest value of hardware engineers: based on technical feasibility, optimizing functionality, enhancing performance, and finding the best balance between cost and performance according to on-site application characteristics; turning a technically feasible solution into a commercially successful product is where a hardware engineer earns their honor.

Some of our hardware engineers may never have thought about where their so-called “R&D design” is reflected, what exactly they researched, developed, or designed? Or did they merely plagiarize something?

The profession of electronic engineer indeed lacks competitive advantages in international competition, and there is a significant gap, which is why there are views that “just copying” is sufficient – your competitiveness is on the level of “just good enough,” and your personal level is also “just good enough,” so of course, both the company and the individual will have a “just good enough” future.

Evaluating whether a problem is high or low is actually related to personal standards. You think this is acceptable, but when you change countries or environments, others might find it inexplicable – how can someone of this level dare to call themselves an engineer? Some emphasize that “such senior, high-level engineers are rare – demand is also low,” which may still be a misunderstanding.

Our current situation is not that there are many mid-to-low-level engineers and few senior engineers, but rather that there are fundamentally few engineers with R&D capabilities; many cannot even be called “R&D.” To put it bluntly, our “senior” might be someone else’s “basic.”

I can’t quite understand how someone who studied calculus, ordinary physics, circuit principles, signals and systems, and analog/digital electronics can graduate and work for a few years and still confidently say they don’t understand analog electronics, don’t know digital electronics, and can’t do small signal analysis… Does this relate to being senior? Just because you’ve reached a certain level can you call yourself a “senior engineer”? Isn’t this the requirement for a basic engineer?

Finally

Remember, the editor once saw a report about a hardware engineer recruitment. The interviewer asked candidates to talk about their experiences in hardware. One candidate pulled out a circuit board from their previous company and said, “Look, it has six layers.”

I’ve encountered some engineers like this; they have very low emotional intelligence, lack sufficient integrity or moral awareness, poor communication skills, and weak learning abilities. What they often say is, “If only there were experts to guide me, and I could work on more projects, my experience would improve.”

They prefer to go on-site repeatedly for debugging (which I see as almost random trial and error; they don’t know why it works when it does, nor why it fails when it doesn’t, and rarely follow theoretical guidance step by step), rather than carefully analyzing things in their heads first – perhaps they genuinely lack the ability to analyze.

Their experience is like a set of martial arts formulas, saying things like “to resist interference, you should ground more points or use single-point grounding,” or “put a few decoupling capacitors in front of the IC,” or “add optical isolation at the communication port,” as if the more formulas they have, the richer their experience.

If you ask them what the circuit principles behind these formulas are, why it must be 0.1uF, and if it’s suitable for this application scenario, they will be left speechless.

In fact, in any profession, the first thing to do is to correct your attitude: do you want to do standard things or “non-standard, just good enough” things? I have always believed that there are huge opportunities in China, and the reason is simple: as long as you meet the basic requirements of being an engineer, you are ahead in the country, with a significant advantage, because there are too many disorganized phenomena among domestic peers or companies.

The problem is, we refer to those who score an average of 90 in exams as “academic elites,” and those engineers who can integrate theoretical knowledge with R&D practice as “experts” (engineers who can’t even combine theory with practice are just causing harm, right?). This is not only a matter of perspective but also an issue of talent supply-side reform: an oversupply of low-end labor resources and a shortage of high-end talent.

This is the talent situation that matches our industrial reality and is also the inevitable path for our future improvement: if the national industry is to upgrade, it absolutely cannot be separated from the upgrade of social talent.

When I communicate with others in the industry, I often say, “Implementing functions in hardware is easy; performance and stability are the hardest.”

I understand that what they call “paper-pasters” refers only to the functional realization of the product. If I give you a product requirement randomly, can you just copy and produce it?

Most of us who have done hardware design should know that those reference designs and DEMO boards, they are not product-level designs; if you copy them directly for use in a product, it absolutely won’t work.

Although “hardware engineer” is a generic term, different engineers may be facing tens of thousands of sub-industries; some work on mobile phones, some on appliances, some on power, some on automotive, some on medical. You could say that any powered physical product has a place for hardware engineers, so many industries, do you think a reference design can cover them all? Each industry’s performance requirements for products are vastly different.

Reference designs are just a user manual for a chip.

A product is not composed of a single chip; a slightly complex product’s circuit board may have dozens of chips, is it that simple to create an overall schematic with just a few reference designs?

Reference designs are a product of the times; today’s chips often have hundreds of pins, and not all datasheets can clearly describe all details.

Even if you are a highly skilled hardware engineer, you still must refer to datasheets, user guides, and reference designs; which engineer can confidently say, “I know everything, I don’t need to look at reference designs?”

I can casually show you one of our schematic design review checklists; don’t you need professional skills to accomplish these tasks?

This is just a part of the design considerations for the schematic of a product in our sub-industry.

Once the PCB design is completed, it requires another round of similar review standards.

A complex chip’s peripheral design is also highly variable; it must meet the needs and compatibility of your overall circuit. For example, the power supply design of the entire machine; you can’t find a complete product-level reference design online because each product has different loads, cooling requirements, structures, safety standards, and cost requirements. The number of considerations you need to account for is simply overwhelming.

I don’t want to say much more. Back to the topic: being a hardware engineer is a high-threshold profession; it’s impossible for you to design independently if you don’t know anything. Such technical opportunists won’t last three days in a company because hardware engineers are strongly interconnected with colleagues from software, structure, systems, procurement, production, and other functions; if someone asks you something, you’ll reveal your lack of knowledge.

To do hardware well, the cycle is longer than software, and once you’ve accumulated a certain amount of experience in a specific industry, you become highly sought after. Hardware is not a profession that relies on academic qualifications; it heavily relies on experience.