Why the Chip Industry Is Struggling to Attract the Next Generation

&quot;Graduate degrees&quot; listed as a reason.<p>Yes, designing chips is hard, it takes a lot of knowledge. This is why medical doctors need to go through all that schooling... designing a tiny chip with more transistors running software that does amazing things is very difficult.<p>My Ph.D. is in computer engineering, specifically VLSI and chip design. This was from a few years ago. I _probably_ should have gone into industry, I mean, after all, it is what I went to school for and wanted to do. However, the starting salary for a chip designer (Intel &#x2F; AMD &#x2F; HP &#x2F; IBM) was literally less than I was making at a side job (I worked my way through my Ph.D) as an IT sysadmin. Not only that, people that I knew well that graduated before me would call me up and tell me it was worse than hell itself. 80 hour weeks? Completely normal, outside of the 2 hours of commute time. Barely make rent because you live in California? Check. Pages &#x2F; Calls all hours of the day outside of work? Check. 80 hours? You mean 100 hours a week leading up to a release, right? Check.<p>Looking back on it, it seems this was &quot;the challenging&quot; and if you made it past this (something like 5 years on) things calmed down for a chip designer and you moved into a more &quot;modest&quot; 60-80 hours a week role with less pressure and somewhat of a pay increase.<p>Yes, how do you attract talent under those conditions? It is not flashy work, takes a lot of schooling and the rewards are low. At least medical doctors can kind of look forward to &quot;well, I can make _real_ money doing this&quot;, and have the satisfaction of &quot;I helped a lot of people&quot;.

The job is too niche. This means there are very few employers worldwide that can make use of your experience. That actually puts downwards pressure on your salary (the employer is a quasi monopsony), not to mention that you are tied to a few locations.

The chip industry obviously isn&#x27;t struggling to attract the next generation, as can be seen by the fact their salaries in the article are about in line with everyone else&#x27;s. I&#x27;ve worked in industries that struggled to attract people, it looked like 2x, 3x, 4x and 5x salaries for what a given skill-set would reasonably be valued at and they still sometimes have to hire unqualified people. And train them, shock &amp; horror.<p>These articles crop up from time to time and I really don&#x27;t think the frame is hitting a high standard of thoughtfulness. Or it is a surreptitious attempt to influence government policy in which case fair enough. The problem with the chip industry is that Asia has a comparative advantage at it, probably because that is where all the advanced industrial capital investment is happening. It has nothing to do with people - it is usual for these matters to turn out to be a regulatory problem when the superficial issues are peeled back.

At least in the UK it just doesn&#x27;t pay enough vs. the effort required, it also seemed like there were also a lot less jobs going compared to software.<p>The article shows some pay figures but those are American where everything pays insanely well compared to here so I&#x27;m not sure how relevant those are.<p>It&#x27;s odd that it doesn&#x27;t discuss the size of the industry though - I always thought of it as a small, relatively niche industry compared to software dev. and while there is probably less competition for that smaller amount of jobs, there&#x27;s still a smaller amount of jobs.

Simpler explanation: it’s grueling work that doesn’t pay market rates.<p>Intel has been the semiconductor industry standard for comp for years, and they’ve lagged in this department for a decade. This has depressed the whole industry’s pay as so many chip companies view Big Blue as the market linchpin to set comp off of.<p>The exception here is for Apple and nVidia - where the work is still pretty grueling, but the pay is excellent.<p>You gotta be the cream of the crop to get in either of those places. The hiring bar for both is high, and getting higher.<p>Source: me. I run a job board for FPGA and RTL engineers.

&gt; The Hardware vs. Software Compensation Myth<p>And then goes to show how a SW engineer makes 30% more. Where’s the myth? Especially given that EE requires a lot more work so that 30% gap is actually worse.

You are too tied to the whims of the capital to risk going down such a specialized field where you need multibillion dollar worth of machinery to get started, I would say.

Theory first<p>The is THE limiting factor of many fields. Bioscience etc. even car mechanics for that matter.<p>It all starts to click when you have a practical and physical application. A garage is readily available and not expensive.<p>A lab on the other hand… imo there should be more low cost options to produce chips. There are a few projects trying to do that. Some universities have some cooperation with fabs for dead&#x2F;spare space.

If they paid new grads $200,000&#x2F;year I can guarantee you that they would not have any issues attracting new talent.

I don&#x27;t understand why they don&#x27;t make certain studies cheaper or free for students that qualify, or even give scholarships. It&#x27;s very important to nudge more students towards politically or economically vital studies.

Hmmm what is a simple &quot;hello world&quot; project in chip design?<p>In computer science courses, that&#x27;s as simple as a println().<p>In machine learning courses, that&#x27;s training on mnist dataset to do character recognition.<p>In electrical engineering, that&#x27;s buying a raspberry pi to blink led.<p>In chip design ... Chatgpt says to design a 1-bit full adder using verilog?<p>...<p>I understand why the article thinks the market is looking for graduate education. To design a simple chip requires an *initial investment* (as with all hardware startups really). This is different from software where one can simply launch a web app with a container hosted on your preferred cloud provider...<p>... That said, with the rise of LLMs lowering the barrier of entry of software even lower (e.g. vibe coding), may we see more rise of hardware startups&#x2F;innovations?

Summary of six key reasons including an insistence on requiring graduate education.

Agree a lot with the article<p>Especially 1 and 3<p>EE education has absolute dog-crap didacticism.<p>EEs also have an awful &quot;holier-than-thou&quot; attitude in Engineering.<p>And then you make it only worse by requiring &quot;Masters only or above&quot;. Well guess why, because your graduation was spent going around stuff that goes from nowhere to nowhere else.