First published on Feburary 26th, 2026
(A peice of writing from Charlie Nicholson)
A collection of thoughts to help people understand hardware, the current state of it, and what we can do to fix it.
In our increasingly digital world, it's easy to forget that every software application, every smartphone, and every smart device depends entirely on hardware — the physical foundation that makes our technological civilization possible. Yet while software engineering has exploded in popularity and investment, hardware engineering faces a critical shortage that threatens our technological future.
This is not a niche concern. It is a national security issue, an economic competitiveness issue, and a civilizational issue. The stakes could not be higher — and almost nobody is talking about it with the urgency it deserves.
This essay explains what hardware is, why it matters, how it differs from software, why we are running out of people who can build it, and what we must do to fix it before it is too late.
Let's start with the basics, because if you work in software, finance, or any non-technical field, you might not actually know what hardware engineers do all day.
Hardware is the physical stuff. It's the chip inside your phone that makes it work. It's the circuit board in your car that controls the engine. It's the power grid that delivers electricity to your home. It's the servers that run the internet. It's the medical devices that monitor your heartbeat in a hospital. It's the satellites that make GPS possible.
While software engineers write code, hardware engineers design and build the physical components that run that code. Think of it this way: software is the instructions, but hardware is the machine that follows them. You can write the most elegant code in the world, but without a processor to run it, memory to store it, and circuits to power it, you have nothing.
Unlike software, hardware cannot be patched with a quick update. If a software engineer makes a mistake, they fix it and push an update in minutes. If a hardware engineer makes a mistake, they may have wasted months of work and hundreds of thousands of dollars in manufacturing costs. The stakes are higher. The timelines are longer. The complexity is immense.
To put it simply: we wouldn't be here without hardware.
Many people, particularly those in management and finance, fundamentally misunderstand what it takes to build hardware. This misunderstanding drives poor investment decisions, unrealistic timelines, and ultimately fewer people choosing the field.
In software development, engineers can make changes and see results within seconds. A web developer modifies a line of code and refreshes their browser. Done.
Hardware development operates on an entirely different timescale. To prototype a single sensor, a hardware engineer must:
Simple changes might take 30 minutes. Complex ones can take days or weeks. When custom circuit boards are involved, iteration cycles can extend to months due to manufacturing lead times.
Hardware projects require extensive upfront research before a single component is purchased. Engineers must understand power requirements, signal integrity, electromagnetic compatibility, thermal management, and manufacturing constraints. A single wrong component choice can compromise an entire system's reliability — or cause it to catch fire.
Unlike software, hardware has hard limits. Components cost money, occupy physical space, consume power, and generate heat. Engineers must constantly balance performance, cost, size, power consumption, and reliability. These constraints don't go away. You can't abstract them.
Hardware exists in the physical world, subject to supply chain disruptions, manufacturing variations, and lead times that have no equivalent in software. The global chip shortage of 2020–2023 demonstrated exactly how vulnerable hardware development is to factors that software simply doesn't face.
[IMAGE SUGGESTION: A side-by-side comparison graphic: "Software iteration cycle" (seconds, free, remote) vs. "Hardware iteration cycle" (days/weeks, expensive, on-site). Clean and visual — makes the point instantly.]
The data tells a brutal story.
From 1997 to 2020, bachelor's and master's degrees across all fields grew by 81.1%. Electrical engineering degrees? Only 37.5% — less than half the rate of growth for all other fields.
[CHART SUGGESTION — Figure 1: Bar chart comparing percentage growth of EE degrees vs. all other degrees from 1997 to 2020. Source: ITIF, 2023.]
It gets worse when you look at market share. In 1997, electrical engineering bachelor's degrees made up 1.1% of all bachelor's degrees. By 2020, that dropped to 0.9%. For master's degrees, the share fell from 1.5% to 1.1%. These sound like small numbers, but they represent thousands of missing engineers every single year — engineers that industry is desperately trying to find.
[CHART SUGGESTION — Figure 2: Line graph showing EE degree market share (as % of all degrees) declining from 1997 to 2020 for both bachelor's and master's levels. Source: ITIF, 2023.]
According to the U.S. Bureau of Labor Statistics, the job market for electrical engineers is growing at roughly double the rate of the average occupation. More jobs, fewer graduates. The gap is widening every year.
[CHART SUGGESTION — Figure 3: Bar or line chart showing EE job growth projections vs. average occupational growth, BLS data. Source: ITIF/BLS, 2023.]
This is confirmed by industry surveys. The Electronic Design 2022 Salary & Career Survey found that 77% of engineering organizations reported difficulty finding qualified engineers.
[CHART SUGGESTION — Figure 4: Bar chart showing percentage of engineering orgs reporting difficulty finding qualified engineers. Source: Electronic Design, 2022.]
The shortage is even more acute in specialized fields. 44% of engineering firms report difficulty finding analog engineers — a subspecialty so demanding it's almost exclusively found in experienced, older candidates, not new graduates.
[CHART SUGGESTION — Figure 5: Bar chart showing which specializations are hardest to hire for (analog, RF, embedded, power, etc.). Source: Electronic Design, 2022.]
The summary is stark: demand is up, supply is down, and the gap is accelerating.
Hardware is hard. It's in the name. But the decline isn't just about difficulty — it's about systemic economic, cultural, and policy failures that we created and can fix.
Picture yourself at 18, choosing a career. You discover:
Path A — Software Engineering:
Path B — Electrical Engineering:
The choice feels obvious — even though both fields matter equally. We have made hardware engineering a bad economic decision for new graduates, and then we act surprised when fewer of them choose it.
Decades ago, U.S. companies moved factories overseas to reduce labor costs. This seemed like smart economics. It was a strategic catastrophe.
Factories don't just need assembly workers. They need electrical engineers to design, manage, and optimize production. As factories moved to China, Taiwan, South Korea, and Vietnam, those countries built massive programs to train exactly those engineers. Today, they produce 10 to 20 times more electrical engineers per year than the United States.
[IMAGE SUGGESTION: World map showing concentration of semiconductor manufacturing and EE talent, with heavy clustering in East Asia. Could show % of global chip production by country.]
The consequence is visible in something as critical as semiconductors. The United States once produced 37% of the world's chips. Today it produces roughly 12%. We depend on Taiwan for the most advanced chips in the world — a single geographic point of vulnerability for our entire digital economy and military capability.
This is not just an economic problem. It is a national security problem.
[CHART SUGGESTION: Pie or bar chart showing global semiconductor manufacturing share by country/region, comparing 1990 vs. today. Source: SIA or CSIS data.]
Software has an enormous startup ecosystem: Y Combinator, Techstars, thousands of angel investors, and venture capital firms willing to back a 23-year-old with a good idea. The feedback loops are fast. The tools are free. The path from idea to product can be weeks.
Hardware has almost none of this. Prototyping is expensive. Manufacturing is complex. Margins are thin. Investors don't understand hardware timelines. There is no "Shark Tank for circuit boards." There is no easy grant pathway. There is no well-worn trail from hardware idea to hardware product without massive upfront capital — capital that almost no young engineer has.
If you are 25 with a great software idea, you can build it at home for free and pitch investors within months. If you are 25 with a great hardware idea, you need $50,000 just to prototype it, and you'll wait a year to find investors who understand what you're building.
The software world runs on open source. Linux, Python, React, TensorFlow — all free. A teenager with a laptop can build world-class software using the exact same tools that Google uses.
Hardware? Almost entirely proprietary. Professional design software costs thousands of dollars per year. Simulation tools are locked behind paywalls. The ecosystem is controlled by a handful of large companies (Cadence, Synopsys, Altium). Open-source alternatives like KiCad exist and are excellent, but they remain the exception. The barrier to entry is dramatically higher than in software — and it keeps talented newcomers out.
Many people entering hardware engineering are surprised by how different the development rhythm is from software. Nobody prepares them. Nobody tells managers. And when timelines slip — because a PCB order takes six weeks — it looks like failure rather than physics.
The cultural mismatch between hardware reality and software expectations creates friction, frustration, and attrition. We lose engineers not just before they enter the field, but after they've already chosen it.
Despite the crisis, the case for hardware has never been stronger. Several massive trends are converging to create extraordinary demand.
We are in an artificial intelligence boom. Every breakthrough — ChatGPT, autonomous vehicles, robotics, scientific computing — depends on cutting-edge hardware: custom AI chips, high-performance GPUs, data centers requiring sophisticated power and cooling systems, and edge devices that run inference locally.
Here is the paradox: while AI threatens to automate many software jobs, it cannot replace hardware engineers. You cannot prompt an AI to design a chip, fabricate a circuit board, or debug a signal integrity issue. Hardware requires physical work in the real world. As AI automates software, the constraint shifts to hardware — and we won't have the people to meet that constraint.
[IMAGE SUGGESTION: Photo of a modern GPU cluster or data center — rows of servers, dense cabling, cooling systems. Caption highlighting the hardware requirements behind AI breakthroughs.]
The global transition to renewable energy, electric vehicles, and smart infrastructure represents one of the largest engineering buildouts in human history. Every EV needs battery management systems. Every solar installation needs power electronics. Every smart grid needs embedded controllers. Every wind turbine needs motor drives.
This is not a five-year trend. It is a fifty-year transformation — and it runs entirely on electrical engineering.
Every connected device, every sensor, every smart appliance, every industrial monitor requires custom hardware design. As the physical world becomes increasingly instrumented and intelligent, the demand for embedded systems engineers will only accelerate.
As long as humanity uses electricity, electrical engineers will be essential. That is not a temporary condition. It is a permanent one. The field offers genuine, durable job security in an increasingly automated economy — and the job market is growing at double the average rate.
We just don't have enough people entering it.
Enough analysis. Here is what concrete action looks like — and much of it can begin now.
Tech companies must raise electrical engineering starting salaries to be competitive with software. They must create faster paths to senior roles. They must offer equity and career advancement that reflect the specialized expertise hardware engineers provide.
This is not charity. Engineers respond to incentives. We gave them every reason to choose software. We need to give them compelling reasons to choose hardware.
The most important intervention is the earliest one. We need Arduino kits, Raspberry Pi boards, and basic electronics curricula in middle schools and high schools — with curriculum that any teacher can run, not just specialist educators. The maker movement has shown this is possible. It needs scale and sustained investment.
[IMAGE SUGGESTION: Students (middle or high school age) gathered around a table with electronics components, breadboards, and microcontrollers — engaged and excited. Shows accessibility and joy of hardware.]
We need hardware-focused incubators and accelerators that understand hardware timelines — programs willing to give founders 18–36 months, not 6. We need venture capital funds staffed by people who understand PCB lead times and manufacturing constraints. We need grant programs specifically designed for hardware prototyping.
A national "Hardware Accelerator" program — analogous to what Y Combinator did for software — could transform the landscape. The startup stories we tell shape who enters a field. Right now, all of our startup mythology is software. That needs to change.
Major tech companies that benefit from open-source software ecosystems — GitHub, cloud providers, large platform companies — should extend equivalent support to open-source hardware. This means funding open-source EDA tools, offering free tool licenses to students and startups, contributing to hardware education resources, and publishing open-source hardware designs where possible.
The democratization of hardware tooling would do for electronics what GitHub did for software: lower the barrier to entry and unlock an enormous pool of latent talent.
Companies that depend on electrical engineers should fund scholarships, endow laboratories at universities, and guarantee internship and hiring pipelines for EE graduates. Universities need resources to expand programs and communicate the genuine excitement and career opportunity in hardware.
We need a cultural shift. Hardware engineers need visible success stories, media coverage, and public recognition. A "30 Under 30" for hardware innovators. Profiles of the engineers behind the technologies everyone uses. A public narrative that makes hardware engineering feel like what it is: one of the most important, creative, and consequential things a person can do with a technical education.
Federal policy should include tax incentives for companies hiring electrical engineers, grants for expanding domestic EE programs, and requirements that federally funded semiconductor manufacturing create pipelines for local engineering talent. The CHIPS Act was a start. It is not enough.
Hardware engineering sits at the foundation of our technological civilization. It is not glamorous in the way software is glamorous. It doesn't generate the same headlines or valuations. But it is the reason any of this works — and we are running out of people who know how to do it.
The actions described here are not aspirational. They are practical. The compensation gap can be closed by companies that choose to close it. The tools can be opened by companies that choose to open them. The startup ecosystem can be built by investors who choose to build it. The students can be reached by educators who get the resources to reach them.
We have talked about this problem for decades while doing very little. Meanwhile, China graduates ten times more electrical engineers per year than the United States. Taiwan controls the chips our military and economy depend on. Our infrastructure ages. Our manufacturing capacity shrinks.
The time for talking is over.
Every device you use, every innovation you benefit from, every AI breakthrough you read about — all of it runs on hardware. All of it needs electrical engineers. The question is whether we act now, while there is still time to reverse the decline, or wait until the shortage becomes a crisis too severe to ignore.
We need more hardware people. We know what to do. Let's start today.
Sources: ITIF, "Short Circuited: Electrical Engineering Degrees in the United States" (2023); Electronic Design Salary & Career Survey (2022); U.S. Bureau of Labor Statistics, Occupational Outlook Handbook; Semiconductor Industry Association.