>>54562213
DBAs are quite highly paid, but it's mostly menial work.

Machine learning and teaching means academia, which is meme-tier. You'll just spend your time crunching out papers about how you used deep learning to make your AI written in Java being able to tell the difference between a black and white zebra and a white and black zebra.

If you're going to do a PhD, at least do something interesting. The biggest buzzword memes right now are deep learning and virtual reality (along with 3D vision I guess).

>>54562259
That was helpful, thanks. I would got for a PhD but that's the problem: I don't find anything particularly awesome to work on, at least for now.

>>54562259
And neural networks, I almost forgot.

>>54562455
Spend your undergrad trying to find out what you like. There is no reason in specialising while doing your bachelors' degree; that's why PhDs and masters' degrees exists.

>>54562213
learn the ELK elastic search stack.
take some db theory.

apply to elastic and make $160k/yr to start and it's not going anywhere anytime soon

I'm thinking about someday becoming a system administrator

>>54562483
I'm nearing the end of my undergrad and I intend to get a master's as I find several areas interesting, that's the hard part. I just meant that going further than that won't happen as I don't see anything worthy of spending my time on a PhD, but that can change with my master's of course.

>>54562511
I'm europoor, would any of this still apply?

>>54562533
From what I've heard sysadmin is increasingly harder to get into

>>54562580
>I'm nearing the end of my undergrad and I intend to get a master's as I find several areas interesting, that's the hard part. I just meant that going further than that won't happen as I don't see anything worthy of spending my time on a PhD, but that can change with my master's of course.
I didn't think I was going to do a PhD either, but I made sure to get good grades so I at least didn't rule that out.

I worked full time during my masters and for half a year after I completed my masters and realised my options was either being a Java/.NET/angular/react/<insert popular framework here> consultant (aka code monkey) or working for complete tools (I ended up trying the latter but hated it), so I went back to do a PhD.

I'm even doing my PhD in something completely different than what I focused on for my master's.

Spend the first semesters doing classes, but look at different theses/dissertations that might look interesting to you.

>I'm europoor, would any of this still apply?
Not him, but elastic search is huge in "big data" area.

>From what I've heard sysadmin is increasingly harder to get into
Not him either, but not in my experience. There is always a need for sys admins. But you will be vastly overqualified with a master's in computer science.

>>54562580
Almost every corp does remote hiring from anywhere these days https://www.elastic.co/about/careers#list

>>54562646
Forgot to add that I'm also a yuropoor and this is my perspective of how it is in Europe (and it corresponds with what I've been told by my international colleagues as well).

>>54562646
>I made sure to get good grades so I at least didn't rule that out.
I'll be focusing on good grades of course.

>consultant (aka code monkey) or working for complete tools
I also do not enjoy code monkeying, which is why databases appeals to me as something where I can still develop and feel much less burnt out, while still enjoying myself. Your latter issue is probably unavoidable, though.

>Spend the first semesters doing classes, but look at different theses/dissertations that might look interesting to you.
Great idea, I should probably start looking into that already.

>>54562659
This does look good. I just don't get how any of these jobs correlate with databases, am I seeing it wrong?

>>54562662
Cool. Which area is you PhD on?

>>54562773
>Cool. Which area is you PhD on?
Distributed processing and high-performance computing.

Or do you want me to be more specific?

>>54562803
I'd like to know more, yeah. High-performance computing in particular is very interesting

>>54562858
Well, I'm working on a project together with a company that makes PCIe adapter cards (so you can make a PCIe-based cluster and have a shared memory architecture). Up to now, I've been working on this device sharing mechanism where a computer can use a PCIe device inside a remote computer but without being aware that it is remote (that is, drivers, OS, application software -- it just works transparently and with minimal performance overhead).

I guess the path forward for that part is to make some sort of clever management/scheduling mechanism on top of that so you can make a sort of resource pool for these devices, where all computers connected together share all their devices and can grab them on demand based on different criteria (for example "I need N SSDs but they have to be close to each other so data transfer will be fast).

I will probably be involved in related stuff, but different things. I'm to have a discussion with my supervisor next week about what to do next, but I _think_ we will decide on making some mechanism for allowing multiple computers to access a disk simultaneously (the stuff I talked about before is still limited by the fact that PCIe devices are really made only to be accessed by one driver at the time), and maybe port one open-source distributed file system to use this multi-disk access thing.

>>54562980
This looks interesting and useful, which is really cool. I can imagine the utility in simply having a lot of hardware piled together and being managed virtually according to the needs of each terminal.

If I had such an idea I'd probably be a lot more interested in getting a PhD, maybe something will pop up as these next two years go by.