PC or Laptop for Grad School?

  • #1


I'll be starting grad school in July and my current PC is a good 3+ years old. I will leave it here so the family can use it, but that means I'll have to buy something for myself.

I don't know if I should buy a laptop or PC, though. I prefer PC's, as I can customize them how I want and when I want. Want a new video card? No problem. Sound card broken? No problem. But, I'm not sure how much more useful a laptop would be, now that I'll actually have real work to do. Being able to have my system with me at all times is something I like. But damn, these things are expensive!

Since I will be in physics, I don't think I need anything for gaming (although I love video games, I am willing to give them up... :( ), just hard drive, memory, and CPU, right?

Well, in any case, I am clueless as to laptops in general, so I was hoping for some advice. Something in the $1k-1.5k range would be ideal. I could spend a bit extra if it was really worth it, though. I'm even willing to buy a Mac! I've been on Linux for a while now, so a FreeBSD variant should be easy to get used to. Unless I can install Linux on a Mac, of course. :)
Computer science news on Phys.org
  • #2
I would think a laptop would make the most sense. I look back on my graduate career and started with a desktop at the house. It was OK for a while then when I started my dissertation work I spent more and more time wishing I had a machine to carry back and forth because at the time I had one at the house and one in my office at the university. It became a real pain having to transfer my code to a disk and (this was before cheap CD's or thumb drives) and carry it home if I was going to work at night on something. Then comes the issue of having mirrored software etc...and invariably I'd leave the latest version someplace and have to go back a day or two.

Now if you want a Linux box, that is fine, but software is an issue, many things take more time to get working software wise. Universities are set up for Windows and I know for a fact that place I went to undergrad has a program to get about anything you need for next to nothing thru the computing services. There is only one or two places I know if to get a dual boot machine all set up which is what I use right now, I work the Linux at home and when I travel for work I use the Windows side.
  • #3
My current school has Red Hat in the phys and astro dept. I actually do what you did right now, except I ssh to and from one computer to another and up/download anything I need. Really stable, since it's a university server that's bound to be backed-up. Of course, I can't be sure my grad school will have Linux on their machines... but I think most everything should be Linux compatible anyway. I mean, it's a science department. At least some professors and grad students would have had to preferred Linux by now, making it more acceptable and having things be made Linux-compatible.

The only software I think I will really need to use is Mathematica/Matlab (like you said, cheap for students), some sort of C/C++ compiler (plenty abound), and some sort of plotting program like gnuplot or root, or even Mathematica. If it's something really bizarre and home-grown, I think I'll still be able to compile it from the source code if I ask the program authors for help.

And hey, worst case scenario is I dual-boot with Vista or XP. No big deal there.
  • #4
Matlab is Linux compatible and you can get any flavor of C/C++ for a linux box. The point I was making about software is that specialty numerical libraries are a little more difficult to get and find in Linux.
  • #5
Ahh, I see. Hmm... I would have thought they'd be on Linux first, but I'm assuming those are proprietary and not open source libraries, yes?

But yeah, I can dual boot with XP or Vista if I have to, so no worries there either. I'm just wondering if it's worth actually having a portable computer if I live on campus and can ssh into the physics dept. servers and store my data there. Like, how often would I need it for things like demos/presentations, or something like a conference or whatnot? I am not planning on doing anything that requires huge collaborations like high energy particles, but I have no idea how life is in grad school, really.
  • #6
If you have the convenience of server space in the physics department, I'd go for the desk top unit especially if you are living on campus. As a beginning grad student I would not figure you'd need a computer for much other than email. In a couple of years I would suspect that you would like to have a laptop. If you are going to do a numerical analysis or computational physics course or two, a laptop might come in handy to work on your code in the library where you have access to reference materials.
  • #7
I see. I'll email and ask if I will get some server space just to make sure. You're right about the reference materials, though. I haven't really had to look up papers so far (they were usually given as handouts or were free online), but I was told that you can only get physical copies at the library if you want them, and you can't take them out of the library, of course. So I'll have to give that some thought... As a physicist, I follow the principle of least action, so the last thing I want to do is something I could have avoided.

I could be more productive on a regular monitor, keyboard, and mouse, but I guess those can be bought and hooked up to the laptop when I'm at home. I've seen it done a few times... And then I could also take it with me when I had to... but that would be the most expensive route. Cheapest would just be a regular desktop that I couldn't take anywhere. I don't know... I'm leaning towards a laptop because of the "just in case" feature of it. I just don't like not being able to upgrade it easily.

Thanks for all your help, by the way. :)
  • #8
WarPhalange said:
You're right about the reference materials, though. I haven't really had to look up papers so far (they were usually given as handouts or were free online), but I was told that you can only get physical copies at the library if you want them, and you can't take them out of the library, of course.

Most journal articles etc.. are available online nowadays: I think I've had to visit the library twice for journal articles in the past few years.

I don't know... I'm leaning towards a laptop because of the "just in case" feature of it.

In my opinion, it's much more convenient to have a laptop. I've been able to work on trains/planes, and whilst away at conferences (even if only to check email in the latter case). I take my laptop with me most days into the office, and only use my desktop (provided by the college) for time-consuming things, or using maple etc. By the way, will you not be given a desktop in your office when you enter grad school?
  • #9
I would go with a laptop. Its smaller for tight spaces, and can be easily moved. I love mine, and I've had it for a few years now. Not too bad of a cost either.
  • #10
Laptop is best, especially since you're prepared to spend $1k. It's more convenient, let's you work anywhere. If you're thinking about a mac, the cheapest macbook is $999. Other kinds of laptops should be cheaper.
  • #11
cristo said:
By the way, will you not be given a desktop in your office when you enter grad school?

Yeah, I'm assuming I will. I guess that comes with server space already, huh? So I will have a desktop to work on in my office, I will live on campus for at least a few years (guaranteed housing), and I'm just trying to figure out if lugging around a laptop is worth it or if it would be enough to have a desktop at home and use the computer provided for me in my office.

Although... being able to have a system with *my* settings and preferences on it at all times would be nice. I don't think they'd let me store my music collection on their servers, or give me root access because I want to install some new fangled IDE I found online or something trivial like that. Or at least, I don't think they give any sort of permissions. I know a post doc I worked with that had a laptop in his office that he just hooked up to a monitor, real keyboard, and mouse. That would work, I suppose.

I still have like 2 months to decide, though. And I definitely won't buy it now, since so many laptops have DDR3 in them, which costs extra, but at this point isn't really any better than DDR2 yet.
  • #12
I think only buying a laptop for $1.5K+ would be a big mistake on your part. For a budget of $1.5k there is no reason that you can't have both. You could build yourself a pretty nice desktop with 22" monitor for $900, and that includes a very nice graphics card, 4GB of RAM, and C2D processor. I can just about guarantee you that the PC your going to have in your office is going to be a piece of @#$% Dell, just like mine so your going to want a laptop as well. Look into the IBM T40-T42 series laptops on ebay. They are cheap (~$175), light (~4.5lbs), and very well built. I used them all through my undergrad and part of my grad as well. They are not exactly power houses but you won't notice a difference between them and a $2k machine if your just using MS office or related programs.

Also, laptops are basically useless when trying to perform any simulations or intensive calcs. It doesn't matter if you spend $1k or $3k on the laptop, they just can't perform as well as a good desktop, even if the desktop has lower specs. Not to mention that having two computers can be very useful if one ever fails.

BTW, Macs are for hippies or people that think they're trendy. They are terrible machines for physicists and engineers, so stay away.
Last edited:
  • #13
Topher925 said:
BTW, Macs are for hippies or people that think they're trendy. They are terrible machines for physicists and engineers, so stay away.

Pff.. nonsense!
  • #14
I have Mathematica installed under Linux at home, and it works great.

I would recommend getting a laptop...perhaps even getting a tablet convertible, if you can afford it. I use mine every day; it's a great way to keep notes and assignments organized (and portable!).

My (physics) department mostly uses Macs, oddly enough. I have no idea why. I usually just use my own computer, because I'm more familiar with it.
  • #15
Topher925 said:
BTW, Macs are for hippies or people that think they're trendy. They are terrible machines for physicists and engineers, so stay away.

That is obvious flamebait... but macs are for people who want the familiar Unix command line with a nice GUI. Half the professors in my computer science department had macbooks, presumably for that reason.
  • #16
Mathematica runs just fine on my Mac. Although, I might consider myself a hippy as well.
  • #17
cristo said:
Pff.. nonsense!

I use VMWare Fusion for any windows stuff I need to run (I use it with a disk image that I suspend for faster access times, and not Bootcamp--though I have that as well for games).

Great when you need to run MATLAB, Word, Excel, etc. (the Windows version) briefly. Almost seamless integration between the two. My school also has MSDN-AA (Academic Alliance) so I was able to get Windows XP Pro (or whatever Vista Business variant is more or less the equivalent) for free, along with Visual Studio.

If you sign up for Apple Developer (Student Edition, something like $20) you can get "development hardware" cheap as well (that's what one of the students in my lab did). You can also get Apple Refurbished hardware (from the Apple Store), which is discounted, or go for the Academic Pricing on new hardware (something like $200). If you wait two months, WWDC (Apple WorldWide Developer Conference) is supposed to bring minor updates (so they'll be clearing out their old stock). You also get cheap AppleCare (the extended warranty) as a student.

They're pricey, yes, but I felt it was worth it (and no, I'm not a clueless 'n00b'--I've been building and repairing computers since the days of the 386). Way more reliable and higher build quality than the cheapo (and yet expensive) eMachines laptop I had.

EDIT: Apple Developer hardware is significantly discounted; Academic and Refurbished, not so much (but still better than nothing).
  • #18
I find a notebook makes the most sense for any sort of academic purpose. The ability to be productive virtually anywhere and at anytime is extremely useful in such a setting. Not to mention the ability to access the Internet (i.e. all the information you could ever need, and more) so long as there is WiFi available, which this day and age is pretty much everywhere, whether it be free or a paid service.

So in short, have a notebook to be your on-the-go productivity workhorse. It doesn't need to be beefy or have the latest chip. On the contrary: pick something light, dimensions no bigger than 14", and good battery life. Remember: you're buying it for productivity and "getting sh*t done", not to play around... that's what your desktop back at home is for. :cool:
  • #19
I personally prefer a desktop due to being able to have a large screen size. I really find myself more productive with a 20 or 22 inch screen as opposed to a 14 or 15. You can, of course, have the best of both worlds and (eventually) buy a separate monitor for use at home.

Also, I've got to mention that having some sort of backup solution in place is vital. I've personally lost two laptop hard drives over the years (even being very careful with them), but never lost a desktop one.

Whether you use a desktop or laptop, my point is the same - ensure you keep copies of things because you don't want to lose your data.
  • #20
Here's what I do: I have a desktop that runs Vista. On it, I have http://www.virtualbox.org/" [Broken] and a virtual Ubuntu machine that is my Linux workhorse. I also have some kernels that I can dissect. For portability, I have a net-book with a copy of Ubuntu on it. The net-book was $300 and does everything I need it to, which is mostly access to the internet, programming small files, playing with root, documentation and as a portable reader. PC, portability and cheaper than a top of the line laptop.
Last edited by a moderator:

Suggested for: PC or Laptop for Grad School?