Entry Level Computational Science Jobs

[jonathan1987]

Ever since I graduated with a Bachelor of Science degree as a Computer Science major, I have been wanting to get into the field of scientific computing/computational science. I have educated myself a lot by taking online courses like Scientific Computing, Computing for Data analysis, Machine Learning, Control of Mobile Robots and Introduction to Engineering Mechanics.

My work background has only been doing research in Astronomy and Robotics and Tutoring at the college level in courses like Algebra, Calculus and Statistics.

What I would really like to know is how come I don't see many entry level Computational Science jobs, when I go to search for them online. Are these entry level positions at academia only or should I do research in the subject in a college/university? (I live in New York City, but all it has to offer are IT jobs, software developer and web development jobs, none which catch my interest).

Any help would be appreciated.

 

[Vanadium 50]

Let's make it concrete. Suppose people are involved in combustion modelling (something computationally intensive and timely). What would you be bringing to that team?

 

[kinkmode]

Maybe I'm missing something here, but most of the people I know who are engaged in computation are not necessarily computer scientists. They are physicists, chemists, applied mathematicians, astrophysicists, or even possibly working in the financial sector. As a result, most of them come from a non computer science background, and most of them got into the work they are doing in graduate school.

Of course, I was in a research environment when I knew these people, so what I witnessed could be a characteristic of where I was. Maybe it doesn't hold true. However, tt makes sense to me, because to write the code properly, you need to understand the science behind it, and most people learn that science in graduate school.

 

[Lavabug]

In the past 2-3 months I have seen a few job adverts for software engineers/computer scientists in big (mostly astrophysical) research institutions like Smithsonian, STSci, NSO, NRAO, and various national labs. These were mostly on job register pages on AAS and APS.

 

[D H]

Forgive me in advance if what I have to say is a bit rough on the edges.

I see a number of problems:

• A bachelors degree as opposed to an advanced degree.
  Yes, people can still get jobs with just a bachelors degree. It's a lot tougher than it used to be, particularly in the field of scientific programming. A masters degree is becoming the entry level degree. The competition is much fiercer for those who's formal education stopped at the bachelors level.
• The degree is in computer science rather than in science, engineering, mathematics, or statistics.
  This is a big negative, at least in my field (aerospace). The computer science majors who do get jobs in aerospace work on the computer science aspects of flight software. Receiving sensor data, sending commands to devices. Making sure all the computational tasks are done on time and as scheduled. Processing the sensor data, generating the commands? Computer scientists don't touch that. They also don't touch the modeling and simulation jobs that represent an even bigger fraction of the scientific programming jobs in aerospace. The industry has been burned too many times hiring a comp sci major who appears qualified, both on paper and in the interview. Resumes from comp sci majors applying for scientific programming jobs are typically discarded.
  
  A comp sci major with just a bachelors degree has about the same chance getting a job as the resume from someone with a degree in philosophy who is currently selling car insurance but has dreamed of being a rocket scientist. The odds that that comp sci major's resume is seriously looked at are in fact less than this nonsense resume. The people who interview (it's a thankless job) need an occasional dose of comic relief. The really good nonsense resumes get circulated. The resumes from a BS comp sci major just get filed circularly (tossed in the trash).
• Online classes.
  Are you doing these for credit, as part of some continuing education, or are you just downloading classes from somewhere and self-educating yourself? Credentials are important in that they are what get you past the first gate.
  
  Get a masters degree in one of the sciences, engineering, mathematics, or statistics and the tables are completely turned. Now that BS in comp sci becomes a huge plus. The above discussion on comp sci majors in general being lousy scientific programmers no longer applies. What does apply is that non comp sci majors in general are bad programmers. People who have demonstrated skills in programming, mathematics, and some science or engineering field are rare. These people are instantly hirable. The good ones put their resume out on the street but once in their entire career. After that the jobs come to them.
• Calling it computational science.
  It's scientific programming, numerical programming, statistical programming, etc., not computational science. The important word comes first. All of these jobs have one thing in common: The ability to program is an essential but nonetheless secondary skill.
• New York City.
  You can find lots of scientific programming jobs in NYC in finance. The pay is out of this world. A low six figure starting salary means you didn't negotiate well. There are also some scientific programming jobs in insurance and in life sciences. Pharma, close. You need to look in Pharm Country (outskirts of the NYC metro area) for that. The rest are mostly elsewhere. You aren't going to find combustion modeling jobs in NYC. The same goes for automotive, aerospace, meteorology, defense. Suburban Washington DC has boatloads of such jobs. So does California. So does Texas. The jobs are spread out across the country.
  
  Be willing to relocate.

 

[jonathan1987]

In terms of online courses, I forgot to mention that I have been taking online courses at coursera.org and plan to enroll in aerodynamics on edx.org. Some of the courses do offer a state of accomplishment for getting a good grade for the course.

For relocation, yeah I have been noticing that many of these scientific computing jobs are outside of New York City, only problem for me is that I don't have enough financially to relocate. But I have a strong hunch that I will have to relocate if I want to be in this field which I do.

I do want to get my masters degree and PhD in either Applied Mathematics, Physics or an Engineering discipline but again its tough financially. At the moment I work part time as a help desk/IT Support at my school making only $11/hr until I find something better.

At the moment I read books in scientific computing and computational physics and just do the exercises in the books and practice my Python, R, MATLAB, C++ and eager to learn other software like ensight, OpenFOAM and other software that these jobs are asking for.

 

[Floid]

Learning is learning, however I don't think many employers are going to place too much weight on online course accomplishments if you try to use that to convince them of your capability in an area.

One option is find a job working non-computational software development with a company that will pay for you to get your Master's degree. That will fix several of your issues and you will also be surprised how much you learn doing real world software development even if it is not in an area you particularly enjoy.

 

[D H]

For relocation, yeah I have been noticing that many of these scientific computing jobs are outside of New York City, only problem for me is that I don't have enough financially to relocate. But I have a strong hunch that I will have to relocate if I want to be in this field which I do.

A halfway decent employer will at least offset those relocation costs somewhat. A really good one will hire and pay movers to pack your stuff up, move it, and unpack it at the destination, *and* they will pay you for your time while moving, 350 miles per day. An even better one will do all that, pay you per diem while moving, and will bump your pay to offset the taxes due on the moving expenses.

Don't worry about relocation costs until you have to face that music.

I do want to get my masters degree and PhD in either Applied Mathematics, Physics or an Engineering discipline but again its tough financially. At the moment I work part time as a help desk/IT Support at my school making only $11/hr until I find something better.

Two comments on this.

One is that you don't want to get a masters *and* a PhD. In the US, its typically one or the other. Most PhD candidates start off as PhD candidates right after getting their bachelors degree. You do not want enter a PhD program because you think it will give you a leg up in some field orthogonal to that PhD program. It's a recipe for disaster. You enter a PhD program because you *need* to live and breathe cutting edge research. A masters program is for those who want a taste of research and who want a leg up in employment outside of academia.

The second item is finances. A PhD program in a technical field should pay you to go to school. You will need to work as a teaching or research assistant, but they will pay you. They won't pay you well (not even close), but they will pay you rather than you paying them. Some schools also offer graduate assistantships for masters programs, but don't count on that.

Another option is to find an employer who will pay your way through a graduate program, typically a masters.

 

[jonathan1987]

One option is find a job working non-computational software development with a company that will pay for you to get your Master's degree. That will fix several of your issues and you will also be surprised how much you learn doing real world software development even if it is not in an area you particularly enjoy.

That is what I have been trying to do. I look for software developer/engineer jobs (entry level) but the requirements for them are so much beyond what they teach you at school. I feel like I have to learn like 5 or more new programming languages/software just to have any chance. Just reading the job requirements for those kinds of jobs makes me feel so negative

 

[kinkmode]

I can't promise there are any real openings (entry level OR in something computer related) but you might want to see what is going on at Princeton Plasma Physics Lab. It's an hour away from NYC by train. They have a lot of people there doing simulation. But they also have a lot of computing resources devoted to other things as well, experiment support, etc. I know there are non-sciencetists who support the science mission. It could be the kind of thing that getting your foot in the door might lead to you getting some experience with the science part and you might also make some contacts with the research people, which would definitely help in terms of getting into a program.

 

[dkotschessaa]

That is what I have been trying to do. I look for software developer/engineer jobs (entry level) but the requirements for them are so much beyond what they teach you at school. I feel like I have to learn like 5 or more new programming languages/software just to have any chance. Just reading the job requirements for those kinds of jobs makes me feel so negative

I think it's also worth noting that these types of jobs are not always "advertised" positions, but the type that are obtained through lots of networking. Looking online and sending out your resume in electronic format is probably not going to work.

By doing I.T. support you are already in the workforce, so that's good. Don't stay there too long. You want to start accelerating. Maybe you can find a similar job to what you are doing but for a different company. Specifically, a company that might do things that you are interested in. Then you're in the door. You might be able to pursue your advanced degree part time. (If you're super lucky you can find a place that will offer partial or full tuition reimbursement).

I just realized kinkmode said something similar to me but more concrete in a way.

Keep in mind that career paths don't usually go in a straight line. Sometimes it can go school-- job --> school --> school/job --> job, etc.

-Dave K

 

[elie_s_dad]

I've done scientific programming in pharma, finance and market research. Most of the points made by 'D H' are not accurate in my personal experience. I know people in aerospace with CS backgrounds as well and really doubt they were evaluated as 'Philosophy car salesmen equivalents' or whatever. They certainly aren't paid as such.

Specifically, at least from what I've seen in pharma, finance and market research:
-CS degree is fine and common. It's also more flexible than most engineering degrees, like Aerospace.
-Employers (decent or not) may or may not pay to relocate. Especially as an entry level person, I wouldn't count on it.
-In the US, many PhD programs grant a Masters. Many people get Masters as preparation for PhD. It is very common to have both a Masters and a PhD. Also in certain engineering fields, universities can fund Masters students.
-An advanced degree has significant opportunity costs (read the rest of the forum). My suggestion would be to gain some experience in order to find the right direction for a Masters, especially if you will be funding it yourself.
-No one cares what you call it - scientific programming, computational science, etc. Employers aren't hung up on labels and neither should you be. If you develop demonstrable skill in those 5+ programming languages (through paid or unpaid internships, academic projects, personal projects, contract jobs etc), you should get some attention from recruiters. Recruiters are looking to put a square block in a square hole - it's easier and more profitable to just play their game than not.

 

[jonathan1987]

elie_s_dad: I actually agree, I mean as long as you know the skills and have a degree in a math,science or engineering degree it shouldn't matter, I believe that the employer should decide whether you are qualified or not.

kinkmode: Thanks for the heads up on the job opening I will take a look at it and contact them to see what they say.

 

[D H]

Specifically, at least from what I've seen in pharma, finance and market research:
-CS degree is fine and common. It's also more flexible than most engineering degrees, like Aerospace.

Your experience differs from mine. That's not surprising. You are working in a different problem domain. I've worked in areas where physics is important: Meteorology, defense, sensor modeling and failure detection, physics engines. With all of the employers I've worked for, we generally did not let CS majors and math mix. There were exceptions of course, the extremely bright person who has proven that we should cast those stereotypes aside.

-In the US, many PhD programs grant a Masters. Many people get Masters as preparation for PhD. It is very common to have both a Masters and a PhD.

You missed my point. In the US, most PhD students enter their PhD program directly after obtaining their bachelor's degree. Yes, those students often are awarded a master's along the way. That's not the point. The point is that entering a masters program as preparatory to entering a PhD program is a non-standard and somewhat wasteful approach.

An exception would be switching to a significantly different area of study, e.g., computer science to physics. Then it does make sense to get that masters degree first. Getting a second bachelors? Yech!

Employers aren't hung up on labels and neither should you be.

That's exactly opposite of my experience. Nowadays, employers and recruiters are excessively hung up on labels. For an employer, hiring the wrong person can be an extremely expensive mistake. Getting rid of that mistaken hire can bump up unemployment insurance and can cause dissent amongst the ranks. For a recruiter, foisting the wrong person on a client can be even more expensive. It might mean losing what was a lucrative client. Employers and recruiters now use mindless filters to reject resumes that don't have the right degree, right keywords. So what if they reject what otherwise have been a good employee? It's much better to eliminate false positives (hiring someone that shouldn't have been hired) even if that means an increase in false negatives (not hiring someone who would have been an asset).

 

[StatGuy2000]

That's exactly opposite of my experience. Nowadays, employers and recruiters are excessively hung up on labels. For an employer, hiring the wrong person can be an extremely expensive mistake. Getting rid of that mistaken hire can bump up unemployment insurance and can cause dissent amongst the ranks. For a recruiter, foisting the wrong person on a client can be even more expensive. It might mean losing what was a lucrative client. Employers and recruiters now use mindless filters to reject resumes that don't have the right degree, right keywords. So what if they reject what otherwise have been a good employee? It's much better to eliminate false positives (hiring someone that shouldn't have been hired) even if that means an increase in false negatives (not hiring someone who would have been an asset).

The consequence of the excessive focus on labels in resumes means that the value of generalist degrees like physics or math are increasingly at a disadvantage for an ever wider range of industries. Would you argue then that if you intend to study in college/university that you should know EXACTLY what field you want to get into, and study that EXACT field, at the exclusion of all else? (For example, wnat to work in engineering - get a BS in electrical engineering; want to work in software - get a BS in CS; forget about physics, math, industrial engineering, etc.; don't even think about literature/pol. sci)

 

[D H]

The consequence of the excessive focus on labels in resumes means that the value of generalist degrees like physics or math are increasingly at a disadvantage for an ever wider range of industries. Would you argue then that if you intend to study in college/university that you should know EXACTLY what field you want to get into, and study that EXACT field, at the exclusion of all else?

Look at all the threads right here at Physics Forums where people with a generalist degree in a technical field complain about not being able to find a job. So, yes, there's a potential issue here.

Those mindless filters that employers and recruiters are so won't to employ compounds the issue.
There are lots of ways to bypass those mindless filters. Go to job fairs. Get a student internship. Become active in a technical society in a field where you want to work. Network, network, network.

There are ways to overcome that lack of specialized training: Get some specialized training. In the US and Canada, undergraduates must take electives to fulfill their graduation requirements. Physics and math students who use these electives to study subjects such as literature, philosophy, or classic languages are aggravating the employability issue. Those electives can be used to ones advantage to make oneself employable. Post graduation, taking a graduate level class is always an option. At many colleges, you don't have to enter a masters or PhD program to take those graduate level classes. They'll gladly take your money.Even though I don't do physics in my job, I'm quite happy to have a degree in physics rather than in some specialized field. It has enabled me to attack a wide range of problems and to be quite flexible in my employment. There are downsides to overspecializing. The field that's hot today will be gone the way of 8 track tapes tomorrow.

 

[zapz]

...You are working in a different problem domain. I've worked in areas where physics is important: Meteorology, defense, sensor modeling and failure detection, physics engines. With all of the employers I've worked for, we generally did not let CS majors and math mix...

Would an applied math major who can program be considered for some of these jobs?

 

[elie_s_dad]

D H - As you say, the differences could be attributed to our different areas. I have only done computational work related to biology, finance and social science. As I said, I do have some friends in Aerospace who are counter-examples to your statements about Aerospace and CS majors. Maybe the exception proves the rule.

Regarding MS / PhD - I don't think I misunderstood your point after reading your subsequent post. My point is that it is common for people to get their Masters before getting their PhD. You call that wasteful and I'd be inclined to agree yet many people do it. One reason could be, as you said, if the person doesn't have the 'right' bachelors to get into his grad program of choice. Another reason could be if the person attended a disadvantaged university as an undergrad from the perspective of graduate admissions (e.g. a low-ranked school or lesser known foreign school).

The OP has a Bachelors in CS and wants to maybe get a PhD in Applied Math, Physics or Engineering. Some CS Bachelors will be prepared for those kinds of programs and some will not; of the second set who are deadset on the PhD, gaining a Masters is a common strategy.

Regarding Labels - You are correct about employers and recruiters and false positive / false negatives and I'm not sure if I argued otherwise. To me, at least in my experience, 'scientific programming' and 'computational science' and etc. are not perceived as meaningfully different from an employers' perspective. Maybe I should have said 'employers aren't hung up on these particular labels because they are not distinguishable signals'.

If a recruiter had two resume, one which identified as 'scientific programmer' and the other identified as 'computational scientist', I doubt he would judge them differently on that basis. Instead he would look to more reliable signals like degrees, universities, previous employer, technical skills, industry specific keywords, etc. His pre-review filters would also be set up around the more reliable signals or they would take both options for the 'bad' signal ('scientific programmer' v 'computational scientist).

At least in my experience, the best way for an entry level person to signal they want to do computational work is to 1) have a technical degree from a reputable place, 2) learn the programming languages used in industry and 3) have proof of 2) on the resume.

A lot of these perceptions could be industry specific.

 

[D H]

As I said, I do have some friends in Aerospace who are counter-examples to your statements about Aerospace and CS majors. Maybe the exception proves the rule.

Of course there are CS majors in the aerospace industry. Just because they are working in aerospace doesn't mean they are doing scientific programming. In general they aren't. Instead, they are working on communications protocols, connecting devices to the avionics bus, ensuring that the flight software will never have real-time overruns, developing a simulation framework, ... There are lots of places in aerospace where computer science expertise is needed.

Regarding MS / PhD - I don't think I misunderstood your point after reading your subsequent post. My point is that it is common for people to get their Masters before getting their PhD.

Yes, it happens. That doesn't mean that it is common. In the US, most PhD graduates enter their PhD program right after getting their bachelor's degree. A small number do enter a PhD program after getting a masters, but it's a small number.

This is getting a bit off topic, though. Except at places that are PhD happy (e.g., JPL, national labs), a masters degree is usually a better option than a PhD if ones goal is to work in industry. The years spent obtaining a PhD are rather expensive in terms of opportunity costs.

 

[elie_s_dad]

1. The people I'm mentioning do do scientific programming. I thought that was implied since it is the topic of this thread.

2. MS before PhD is a common track in my experience.

3. This is getting off-topic. The point I was making that my experience does not agree with your observations. That could be because what you were saying was Aerospace industry specific.

 

[jonathan1987]

Hey Guys sorry to reply months back later, but I was wondering, should I get a job working as a entry level software developer/engineer and then after a year or two go into scientific computing just so I can have some experience? I'm not really an IT guy, web development or mobile development kind of person.

 

[jesse73]

If a recruiter had two resume, one which identified as 'scientific programmer' and the other identified as 'computational scientist',

Thats naive. The use of software and search programs recruiting has made it so that a lot of "keyword" searching is done by the recruiters so that what you label yourself matters.

 

[SophusLies]

Ever since I graduated with a Bachelor of Science degree as a Computer Science major, I have been wanting to get into the field of scientific computing/computational science.

What I would really like to know is how come I don't see many entry level Computational Science jobs, when I go to search for them online. Are these entry level positions at academia only or should I do research in the subject in a college/university? (I live in New York City, but all it has to offer are IT jobs, software developer and web development jobs, none which catch my interest).

Any help would be appreciated.

My first real job was with a computational physics consulting firm. I only had a bachelors in physics and math with a mess of CS classes at the time. From what I've seen, DH is right in terms of work done by actual CS degree holders. Even with my background, alas bachelors only, most of my job was debugging and testing code (C++). If there a memory leak issue, I was asked to figure it out. Any of the real computational work was done by master's degrees or PhD's in either physics or engineering or even statistics; this involved coming up with equations for the models, limitations of the models, and how to numerically solve them.

I would recommend looking at financial firms and maybe try to get into that field. I recently was in NYC over the summer doing an internship (I'm a PhD student currently) at a financial firm. Those places pay a lot of money and the work is very interesting. I think anyone interested in a heavy math type area would love financial modeling. It involves so many different areas of math and cross-lines with lots of CS stuff too. I doubt with only a CS degree you'll be doing much math but at least you'll be close to the stuff and at the very least you'll have to learn the code that crunches all the numbers. :)