<p>I am a junior and will be doing stuff like CAD, c++ and matlab. i recently bought a macbook air 13 inch with 4gb ram, 128gb ssd, and i5 1.3ghz processor. Its pretty sweet and it does run cad and everything. However, i want to learn solidworks as my professor said having it on your resume helps with getting an internship. Should i get the 2012 macbook pro baseline and upgrade the ram to 8gb and eventually the hard drive to ssd? I really like the air but everyone is telling me t hat 4gb of ram isnt enough and it wont last with all the new software updates. i plan on having the laptop for around 3 years hopefully more</p>
<p>A few things:</p>
<p>As a general rule, engineering does not work well with Apple. Apple makes and sells trendy consumer products meant for relatively light processing loads largely confined to multimedia applications. Engineering needs workhorses that can handle often tremendous mathematical loads.</p>
<p>Engineering also operates in opposition to trends in consumer computing in general. Most people want ever smaller, lighter computers, but engineers see the RAM and processing power required increasing every year, often faster than actual improvements. This means engineers get LARGER computers, not smaller. I know you love the air, but take the RAM instead.</p>
<p>Also bear in mind that soon this will not matter. If you are a junior not going to grad school, in a couple of years you will be in industry. They will provide a computer for you, and unless you pick a very specific kind of company it will NOT be an Apple product. Among other issues, Apple is so consumer driven that it leaves out a lot of features (security, mostly) that corporations want to see. So in 3 years your MacBook and iPhone will be strictly personal, and will set next to the BlackBerry and IBM EliteBook that you do all your actual work on.</p>
<p>SolidWorks does not run on Mac OS. You would have to run it from a Windows installation using Boot Camp or some other similar solution. This means you will have to dual boot, meaning you will need multiple partitions and two full file systems for your two operating systems in order to run SolidWorks, meaning you are really going to be stretching that 128 GB of storage that you have. You can overcome that mostly by intelligent partitioning though.</p>
<p>4 GB will be enough RAM for most consumer computers for the foreseeable future. The question then is how much of that heavy computing do you actually expect to do on a laptop? Unless you plan to get into reasonably complex CAD assemblies, scientific computing, video editing, etc. on your laptop, then you probably don’t need to spring for too powerful of a system. That said, RAM is cheap, so going to 8 GB is certainly advisable if feasible.</p>
<p>
</p>
<p>I’d argue this is true if you are using commercial software. However, I know a fair number of computational engineers who use Macs religiously, especially the Mac Pros, which were absolute number crunching machines. Of course, most of them seem to have wiped OSX off and installed Linux for real, heavy numerical work.</p>
<p>
</p>
<p>I’m sorry, but the part about security is just wrong. If companies like Google are okay with OS X’s security features (this is the same company that uses self-destructing flash drives), then there is nothing to worry about. I can name plenty of other corporations (mostly tech companies, since that’s my field) that have widespread Mac installations.</p>
<hr>
<p>I think it makes more sense nowadays to have high-performance servers doing most of the number-crunching. Better yet, clusters.</p>
<p>
Security is in fact the exact reason why my company does not use them - the group that was evaluating them (using iPhones instead of BlackBerrys and macs instead of PC’s) were disappointed not only that they had to give them up but also how many of the security tests they failed. Contacting Apple confirmed that it was not possible to install the required capability even for a purchase involving thousands of units. </p>
<p>Security concerns and methods are not uniform across industries. Just because Google uses them does not mean that everyone can or should.</p>
<p>
Yes and no. “Routine” engineering work is getting intensive enough computationally that you need to either need a powerful computer or need near-continual access to servers or clusters. Both are problematic in one way or the other, but for most people it is not worthwhile to sign onto a cluster every time you run a simulation. So you get powerful computers that can handle 90% of the work at the individual level, and then send the top 10% of jobs to the cluster.</p>
<p>FWIW, I use my work computer for simulations and calculations that take up to 8 hours to run. More than that and I usually do send it to the cluster.</p>
<p>
</p>
<p>It may take only 1 hour if you send it to the cluster.</p>
<p>
Indeed, which is the point… but arranging for it, waiting for your turn in the queue, waiting for it to process, and the relative costs of the cluster itself can make it a bad deal for routine jobs.</p>
<p>There is also the fact that depending on work conditions, you may not always have access to the cluster, and that for security concerns you may want to keep the data off of something with so much access.</p>
<p>Hmmm… I think we are getting a bit off topic. This thread is not about cluster vs desktop.</p>
<p>I took solidworks last year…The professor only had bad things to say about the Mac. Honestly just about every one of my engineering professors has talked about its hinderence in engineering. I just finished this advanced statistics course and apparently the excel that comes with Mac doesn’t even have all of the same abilities to do data analysis.</p>
<p>
</p>
<p>Fair enough, I suppose. I guess I just wanted to point out that not all corporations are against the OS X/iOS platform for security reasons (i.e. they have done their own security audits and come to a positive conclusion).</p>
<p>And yeah, I imagine the cluster situation is also somewhat situation-dependent. The company I worked at previously had a huge Hadoop cluster running 24/7 (so will the one I am about to join). So it was very routine to run jobs whenever you wanted for whatever reasons.</p>
<p>Anyways, back on topic. OP, you can keep the Air and dual-boot with Windows if necessary (it’s good enough for the next two years), but as has been mentioned, whatever company you join will provide you with the necessary equipment.</p>
<p>Most HPC solutions (clusters an full supercomputers) have such a backlog of users submitting jobs for processing that for “routine” engineering work, the time spent queuing for processor time negates the benefits of HPC. For those situations (which are many), it is nice to have a personal machine that can do relatively simple computations in the same time it could be done including the queue for a cluster. Where HPC machines really pay off is when your problems are so large that the time savings derived from the HPC are much greater than the time spent queuing for processor time.</p>
<p>Here’s my 2cents, your school is going to have computers you can use to do what work you have to do at the school, it will most likely be windows based. If this is true, you should have a windows based computer at home because that way any work you started at school you can finish at home or vice versa. Now, knowing that engineers use computers so much in their jobs, and that most of the industry uses windows based machines; you should seize this opportunity and make your own windows based PC, and not only have a computer customized to your own needs but you’ll have learned a great deal about computers in general.</p>
<p>my opinion, spend money on a mid level motherboard, preferably an Intel LG 1155 based one, that can support at least 32gb of 1600mhz ram. Buy an i7 4000k or whatever the k model for the i7 is, I can’t remember the exact name, you want the k model of i7 BC its one of the fastest, has hyperthreading, and the k in the name means you can overclock it. I wouldn’t spend a lot on a gpu, buy an entry level, just so you have something better than integrated chipset, you don’t need a powerful gpu as you won’t be gaming or doing video rending/photoshop which do require a great deal of vram. For storage I would buy an ssd for boot up of windows and application then a 7200 rmp 64 mb cached 1-2tb hard drive for all other data(preferably a wd black enterprise edition). The size of ssd should be at least 128gb with 256 GB at max, you won’t fill that.</p>
<p>once you bought and made the computer research ssd optimization and how to keep your data nice and organized by utilizing ssd/hdd combo. Then read up on over clocking CPU and what kind of cooling you’ll need.</p>
<p>I forgot to mention for the price of what you are going to spend on a mbp with those upgrades, retina or not. You can make a killer PC and do everything I said and more or just what said and save a great deal of money</p>
<p>^ He wants a laptop not a desktop.</p>
<p>No, he never specified he wanted a laptop, he said he has a mac air and is asking if he should (I’m guessing this part) sell the air and buy the macbook pro, which he have for ~3 years. I’m suggesting an alternate route in that saying no, make a desktop, the school should have its own computers you can use anyway.</p>
<p>
He will already have access to desktop computers and currently has and is seeking laptops. Desktop advice is probably not that useful.</p>
<p>“he will already have access to desktop computers” I’m assuming you mean at school and that’s entirely my point. He can use the schools computers at school, so what does he need a mobile computer device for? Therefore for home use I suggest a desktop: better performance, costs less, highly customizable, and will gain a lot of knowledge in building one</p>
<p>And his point is that he can already use a workstation-class desktop at school so why not something portable that he can take home, to the library, to presentations, etc. That’s the reasoning behind getting a laptop.</p>
<p>The thing is, people shouldn’t be trying to buy a laptop based on the performance necessary to run CAD and heavy computations. You end up negating the whole advantage of a laptop.</p>