SAN JOSE, Calif. – Systems engineers are a vanishing breed at Hewlett-Packard. I suspect that’s the case for many OEMs, and I’m concerned what that means.
Earlier this year, my company was doing a series of interviews with engineers under 30. I asked a long time HP contact if he could put me in touch with a couple young EEs.
His response was like a splash of cold water. HP hasn’t hired any young engineers in years, he said. The company has focused on retaining a core cadre of veteran engineers with deep expertise; many younger engineers fell victim to the various waves of layoffs over the last few years, he said.
Another long time HP contact fleshed out those observations over lunch at an industry conference recently. Only a few computer systems engineers are left at HP, he said. They work closely with HP’s contract manufacturing partners, typically in Asia, overseeing their work, he said.
These systems engineers are approaching retirement age. When they are gone, there will be no one left to replace them, he added.
HP is not alone, but its situation is not exactly universal either. Recently, I was talking with a veteran IBM ThinkPad engineering manager who now works for China’s Lenovo, the smartphone and PC maker that acquired IBM’s notebook group years ago. I asked him about system design at Lenovo.
“We are unique in this area because we do our own design work and manufacture quite a bit of our systems, too,” he said. “Some systems we have made by ODMs, but even then we do the design work ourselves--that’s one of our key strengths,” he added.
I’m sure IBM still has several systems design engineers creating those custom boards needed for its mainframes and Power servers. I don’t know if it actually makes any of the boards, but I doubt it.
Last week, Applied Micro Circuits Corp. showed four dense, complex server boards it created as reference designs for its X-Gene ARM server SoC. It designed one of them with engineers at Dell, I suspect some of the old hands there that—like HP—don’t do much design anymore but tend to specify things and work with ODMs in Asia.
There are young computer systems engineers in Silicon Valley. I met four of them recently. They work for Facebook, designing what goes into its data centers. They have young systems engineering peers inside Amazon, Google and Microsoft—but not Dell, HP or IBM.
So it goes.
The issue spans notebooks and desktops as well as servers. I know electronics design has become relatively straightforward for many Wintel computers over the last 20 years with more focus on industrial design and software than chip and board-level choices. But the winds are changing.
Waves of x86 and ARM SoCs from Intel, AMD and a whole range of new ARM licensees are coming down the pike between now and 2014. Right behind them is another wave of SoCs using 3-D stacking that will add a whole new level—as it were—to the technical choices.
I think OEMs are going to have a painful need for systems engineers over the next couple years. Good luck stealing them away from Google and Facebook!
Will HP really need systems engineers?
Old timers may recall when HP moved away from their Precision Architecture RISC server line to standardize on Intel.
And HP's move back then was simply part of a progression that had already been occurring. The underlying problem is that R&D to develop and maintain custom solutions is fantastically expensive, and it gets harder and harder to generate an ROI to justify it. Increasingly, if things *can* be done with off the shelf components, they are, because it's cheaper.
The model for the future might just be Visio, the TV maker who recently expanded ito PC products. Visio outsourced a good deal of the systems design to its manufacturing partners, and spent a lot of time talking to Intel about hardware and Microsoft about software before proceeding. They told their partners "These are the products we want to make, and the price points we want to hit. You tell us how this should be done."
Visio is essentially a marketer and packager. Their design efforts are concentrated on what the product looks like to the buyer, not on what's under the hood.
HP wants to emulate IBM and move more towards a software and services model, and while they have decided to retain their PC operations, they are looking at ways to move up the value chain and provide products for which they can charge a higher price and make a better margin. I'm willing to bet the peiople they will want will be industrial designers doing packaging, and not systems engineers. Systems engineering can be done by people at HP's manufacturing partners, with HP providing specs for what they want the design to do. HP might have an experienced system engineer or two helping to develop the specs, simply to insure the specs are reasonable and can be met by the partners at the desired cost levels, but they won't need to do the detailed design.
Given the direction HP is going (and the industry as a whole is going), why *will* they need systems engineers?
That's probably one of the reasons why HP PC's are cheap and have poor quality. I bought an inexpensive HP desktop in 2008. Within a year, the SD card reader stopped working. Then the DVD player/burner went bad. Then the hard-drive connector became loose. I'm still using it, but have to keep the computer case open, so I can re-plug the connector now and then. I am very, very, disappointed with the HP brand.
Why can't HP produce high-quality PC's, and sell them at a premium price? It can't win the race to the bottom.
Doesn't take a genius to see the writing on the wall: consumer systems are going towards lower performance very integrated SIP-based designs. When the current generation of HP systems engineers retire, they aren't going to be replaced.
These are sad days. I'm 49( feel like 22 ). I grew up in a small rural town. Fairly poor. Self taught in electronics until I went to college for EE then onto Qualcomm,Microchip,Hypercom,Sicom/Intersil, now gov.contract R&D. The ability to be versatile and do HS/SW/Embedded/Machining is gained through experience. Most kids don't seem to have that homegrown hands-on attitude( excepting the Arduino prodigy's :-)many 3rd world country kids have to create just to get access to engineering and it puts them ahead. Playing in all these fields is what leads to the years of experience gained by these older system designers. I often see management cutting to the bone and hacking out these guys for newer grads, so it's nice to hear HP is forward thinking not only of profit, but quality and longevity of products. Sometimes pride and quality of product needs to be put ahead of bean counting, otherwise planes fall out of the sky. OEM.
Guess that makes me one of the REAL dinosaurs. If one were to restrict the definition of "systems engineer" to someone who writes the system specification in a purpose-built tool like DOORS, then follows through to verify that those specs are correctly implemented through testing, there are few of us left indeed. I guess I "cut my teeth" in the Reagan defense buildup that ended the Cold War, some of us were "hiding out" in commercial avionics because it's just about impossible to get a design certified through the higher levels of DO-178 without us. Problem is there isn't much going on in the latter field between the horrible economy and the president badmouthing the people who use business jets. It's my opinion that it's mainly in safety-critical fields certified to a written standard that system engineers are correctly perceived to earn their keep, since in more competitive environments management still gets to retain the prerogative of changing nearly all the specs the day before delivery and cite "competitive pressures". Do that on a 3-5 year DO-178 project and you'll lose both your shirt AND your customer!
This article is very inaccurate. I know an engineering manager in hardware at HP that actually has the exact opposite problem: they are actively (and have been) hiring college hires for years, and they are missing experienced engineers for mentoring and other work. I think he takes a very small view of a company and uses it for the authors point.
And no, I am not employed at HP not do I desire to be.
What do Facebook, Amazon, and Google have that Microsoft, Dell, HP, IBM don't have that attracts these systems engineers? Upside potential in financial compensation (salary, bonus, stock incentives). They also have less "paleo" cultures as they're not old, stogy, mature companies that younger engineers don't want to work for. The culture's a bad fit.
What kind of education background is needed for a systems engineer today? A BS CPE probably won't even be enough to get you an entry level position pretty soon as you just know the basics but it'd be 3-5 years before they'll have trained you to really get nominal productivity from you.
I suspect you'll need a MS degree in some specialized area they're looking to fill to get hired. By then you'll be an indentured servant to $150k in student loan debt, so the only way out of that is the upside potential of sweat equity in a fast growing business.
Based on a tour of the Menlo Park HQ, the Facebook environment is much more appealing to young people than say HP and Dell. It's a happening place!
As is the Apple campus, too, where I was lucky enough to have lunch today.
Someone who wishes to remain anonymous emailed me to say:
"In San Diego, HP has been laying off and offering early retirement to experienced firmware engineers and then immediately turning around and replacing them with new grads. They have been doing this for some time (several years, at least.)"
This is another reason we need industry leaders that can think beyond the next profit/loss statement. When I worked at 3M we had an incredible program to bring young engineers in and teach them the ropes of actually making things, process control and "common sense". It is not cheap to make quality engineer - or an worker.
You long term survival depends on nurturing these long term and has to depend not only on companies with some long term views, but a society and government to give the basic education needed to get potential engineers to the next level.
This is such a difficult subject on so many levels and deserves and entire section here as without Engineers - trained engineers being hired and trained consistently we will not be competitive. As a person who worked myself up from military training in electronics (1 year in school analog to digital - we now use contractors and let most military switch modules...) and working as a technician to Engineering Tech to Engineer I just don't see that development path as available today - back in the 90's they were removing Tech's from Engineering labs - how do you work your way up? How many of our great engineers had it all together and had the grades, money and desire to get to a four year school, and then the stamina to complete and engineering program the "normal" way. Engineering takes creativity and technical ability and in my opinion the environment we currently have makes it MUCH more difficult and many of our greatest engineer would not even have been able to do the incredible things they were able to do over the last 40 years - one of the things that made America great. A mind is a terrible thing to waste - why are we making it harder? Although I do have to say with the internet and such it is much easier to get information than in the 70's and I very jealous of today young enthusiasts.
Facebook isn't very interesting unless you want to work on networking hardware... Which is why I like this article.
I think the biggest thing with getting new EEs, is getting people into the engineering mindset... It's a mix of creativity and practicality that comes from being cross-discipline. Very few young people know how to sell themselves and create their own market/product ideas...There's a big reason for this.
If you don't use science as your foundations, and you don't see that everything you learn is not useless, you just won't enjoy engineering...Once you can see it in everything... It turns into your life, and you can basically FOCUS on "adding" to what we have, without the typical failures of unrealistic goals... I mean I could even reverse engineer some ridiculous consumer item and think of an accessory, then form a business model for just selling that...Young people are just like "oh my parents paid for this education! It's free!" ...Well they are not going to understand leadership at all with that mindset...It's just hard to inch people out of the comfort to show them that the discomfort produces a temporary result.
Life is not ideal, people don't understand this any more... It's so hard to get it to appeal to people, because they think more about the reward, than how we all started... If you turn time back to the era the older people here were born in ( and I'm only 27 )... Quality was understood as relational to price and man hours... Now it's not that important (which is wrong) since we tend to pretend we live in a world where technology has solved our problems... Which hilariously murphy's law destroys this all the time.
The biggest issue I have with electronics theory and teaching it to anyone... Is getting people to see that the more developed a technology becomes...Making it still takes more effort... You have to see both sides, there's never an argument that something is bad. It's good at one thing and bad at another... It's everywhere, you can't use a single thing that isn't electrical now. Once you can teach people that and say "THIS IS GOOD." and that you're never going to go anywhere in life without using a QUANTIFIABLE amount of effort... It's the value system we have in place. Most people leave because of what technology they value and jump fences... People seem to think we have broken out of the old laws of physics and thermodynamics, so they just skip it... It's because of media and modern social influences. It's not exactly bad, but it's created rifts in our economy. But to anyone reading this if you are smart, this means you can fill those gaps.
I really think everyone can learn the engineering mindset...The main thing about it is learning that you can always improve a design, and that something was designed by a person. I will show someone a piece of hardware, and then I often ask them how they feel... Then I explain to them how they would design it in their current situation in life.. Lots of companies just started out in a garage I believe. Success for young EEs should be based on willingness to always ignore failure ( consider it a constructive part of the process ) and try something new. It's a trick I learned from the old dogs... :)
Electronics corporations are often responsible in letting thousands of workers go in downsizing operations. Look at Nokia, HP, Cisco, ... the list is long, and we assume those let go did more than make tea, or watering the regulatory pot plants setup by landscaping companies in far better vehicles than the engineers. One of the biggest laughs is the "test and tag" trend in checking mains cables periodically without a clue of what they are doing. They will happily attach anything to their tester, but cannot tell you what test voltage will be applied. Nice cars as well! Would not recommend this as a career to kids.
Herb Simon, at Carnegie Tech in 1950's, told us that automation would eliminate most jobs eventually, but the last to go would NOT be we engineers, who will be easily replaced by computerization, but the bulldozer operator with his hand-eye coordination talent, something the automation gurus had a great deal of difficulty replacing. The net result in our time, is that a very small number of very talented systems thinking can use much better subsystems, now that standards have evolved, to fill the work needed....or so it seems.