The original samsung series 9 is a sandybridge "ultrabook" from 2011 that can take 16GB of ram. While technically before ultrabooks were first released, it has a similar profile to the air. It was also released before 8GB dimm modules were available, but while never being "supported", they work just fine.
I haven't seen an ultrabook I'd prefer over it, especially when I rule out ones that use a tapered design.
My macbook pro isn't that much thicker than an air.. though I wound up getting the 15" retina because I needed 16gb without waiting a few weeks for it... tried an 8gb/13" retina model, and returned it, needed the extra ram.
If you want more memory you have to go to the T4XXp series, and probably buy the RAM yourself as the premium charged is considerable: 16GB on NewEgg is $120, while the premium payed for 16GB out of the box is $340. The T440s also maxes out at 12GB.
I did this. I had an X1 Carbon, but couldn't find a way to upgrade the SSD (it's some weird size and connector) and 8GB wasn't enough, so I sold it and bought a T430s.
I'm curious, but why do you need more than 8Gb for ? I can think of a few potential uses of more RAM (large video files edition, R on very, very big data sets) but 8Gb seems reasonable for most tasks. Are you really in a situation where you need 16Gb every single day ?
Are you really in a situation where you need 16Gb every single day?
Maybe not every single day, but easily 3 days a week. Personally I could never consider owning a primary work machine with less than 32 GB of RAM, as that is just about enough (with some swap) to do something reasonable with a moderately sized data set in reasonable time, without having to try to get too clever. That being said, I currently have an MBA with 4 GB of RAM as a secondary computer to complement my workstation at work, and that's fine for most of my day to day programming tasks.
> Personally I could never consider owning a primary work machine with less than 32 GB of RAM, as that is just about enough (with some swap) to do something reasonable with a moderately sized data set in reasonable time
Suffice to say, you probably don't represent the average developer, or power-user. I can write code just fine without having to resort to 32GB data-sets. I think most other developers can too.
Not saying I don't acknowledge that some people may have such a need, but I don't see this is a big enough issue for enough people, to think it warrants making the already expensive developer edition even more expensive.
If you need 3-digit Gigabytes of RAM, have you considered just remoting into a server which has all the juice you need instead? In the age of the cloud, why on earth do you need all that power in your laptop? That just seems awfully backwards.
32GB might be somewhat excessive for mainstream developer laptops, but 16GB is not. A pair of VMs (for various MSIE versions, isolated dev environments, ...), an emulator or two (Android, iOS, …), 2-3 browsers open (with many tabs in at least one of them) and an IDE and you've blown way above 8GB working set without working on any big dataset.
I'm currently on an 8GB machine and regularly have to pare down my working set to avoid swapping.
I don't see this is a big enough issue for enough people, to think it warrants making the already expensive developer edition even more expensive.
Sure, I agree. And I'm sure I'd be happy with this machine as secondary coding machine. But I do dream of one day getting a single laptop that covers all my needs.
have you considered just remoting into a server which has all the juice you need instead?
I do the few times I need 100+ GB of RAM, and to be honest it's a bit of a faff. Getting your tool chain set up, copying huge data sets back and forth, working out how to install proprietary software and the licensing there of. I'm unbelievably ecstatic that the option exists when I need it, but I'm equally glad every time I don't have to deal with it. Most of time my data sets aren't that big and all my work fits comfortably in 64 GB of RAM.
Yeah, I also work quite a lot with large data sets, but what I want from a laptop is for it to be really great at doing everything else. I'll ssh to a machine that cost more than my car when heavy lifting is required.
I had to sit on my hands to keep from giving Dell my credit card number reading that review.
An Java app, IntelliJ IDEA, few Docker VMs, 2 browsers (and you know how much memory Chrome eats?), few not so complex tools for development (like SourceTree, Dash, etc) - and 14Gb are used. Just checked. And I didn't run anything serious yet.
Had 8gb before, machine was swapping too much, become unresponsive, etc.
I run into the 8GB limit on my laptop pretty regularly. Mostly due to VMs for testing and building for other systems. But I also recently found that running Android Studio, plus the emulator, plus my usual stuff pretty much shuts my machine down. I had to use a VT to kill my mail client, the VMs, and some other processes, in order to get the mouse and keyboard responding in X again. Java memory usage is kinda crazy.
This is my use case for having 16GB on my laptop as well. Being able to simulate the network architecture of our infrastructure with a bunch of little 512MB-1GB VMs is a nice thing to be able to do.
If I'm not doing development with a cluster of VMs though, I rarely go over 8GB.
I use a Dell E6420, as a data scientist I kept breaking the default 8GB so upgraded to 16GB which is almost always fine. I won't go >8GB most days but when I do, I need it (else I lose hours trying to partition data and think about ways to subselect). 16GB for data science seems a sensible minimum for the folk I know. I'm on Linux Mint 17 (Ubuntu 14.04) + Python.
I work in games industry(as a programmer) and my machine has 64GB of ram. I hit swap pretty much every day, and our IT is currently in process of upgrading all machines to 128GB. In my own time I develop games in Unity and I used to run out of ram on an 8GB machine - upgrade to 16GB was an absolute necessity. I would not even consider buying a laptop with 8GB nowadays.
I have 16GB in my work laptop, and I find myself wishing for more constantly. I support a Linux application that wants 24 GB of RAM for itself. I'm also client-facing, so I have a Windows VM in order to run Microsoft Office. 16GB of RAM is plenty to run RHEL and Windows, but not quite enough to run RHEL, Windows, and a VM of the software I support.
Indeed I am, my "minimum useful" amount is 16GB -- my desktops are 128GB, and my current "monster laptop" is 64GB. The in-flight state on the software I work on can get well over 8GB with our test dataset.
I have 8gb and get warnings quite often that I'm running out of memory: windows 7, ubuntu vm with oracle or Postgres, IntelliJ, Firefox, chrome, office, R. Both R and the virtual machine take 2gb+, IntelliJ 1gb+...
Well one plus is the total RAM footprint of a Linux distro + a DE, is less (by a lot) than just OS X's kernel. Right now, xnu alone is using 960MB on my laptop. Oink oink oink.
Edit:MBP82, created new user accounts, reboot, login, launch Firefox, load CNN, wait 2 minutes. Did this twice each Fedora 21 Workstation (Gnome) and OS X 10.9.5. Fedora free -m used = 614MB. OS X top PhysMem Used = 2866M, Activity Monitor Memory Used = 3.69GB. I suspect Activity Monitor includes "wired" which top splits out. shrug Anyway, it's not a tiny difference. Omnomnomnom.
The RAM footprints of most development tools on Linux are also less than on other operating systems, due to shared libraries and not just shipping every dependency bundled with binaries. A developer on OSX told me he was close to maxing out his 16gigs of RAM, blaming it on having to run 2 vagrant VMs simultaneously while working. While I wasn't able to establish a conclusive cause of his sizeable memory consumption (he was busy), I spun up 3 vagrant VMs of my own, simulated work, and opened/visited many more sites in Firefox to try to hog resources. I could hardly break 4 gigs of RAM used (out of 16), in Ubuntu 14.04. 8 gigs might not actually be so bad if you're doing development work using tools that aren't shipped as complete packages, but rather adhere to the unix philosophy.
I suspect a lot of this comes from OS X apps all shipping with most of the library dependencies, so there's no way of deduplicating libraries that are loaded in the memory. In Linux, shared libraries are used, at least with software installed by package management. OS X should benefit from a kernel feature something like Linux KSM (kernel samepage merging[1]), that scans for duplicate memory pages and merges them to shallow copy-on-write clones.
One can hope, currently as my next portable, I am looking at the Asus UX303LN (great memorable Asus names...) but haven't found much info on putting Linux on it yet.
16 GB is a must for my next machine, nothing I do is that CPU intensive anymore.
I'd also like a Dockingport... its the main reason i love my Surface Pro 2 so much.
Displayport, Dockingstation, 8GB Ram and 512GB SSD all below 1200g, that is to be beat. I hope for a 16GB Ram version for the next Surface Pro 4.
I run daily into swap, having just one VM open, Photoshop and PHPStorm...
Hold off a bit then. USB 3.1 Type C is just around the corner. Once it's established, pretty much every laptop, tablet and phone will contain a USB 3.1 Type C docking port. (and most monitors will be docking stations).
Do you have a particular timeframe on that? I'm in the market for a new laptop, and considering just getting an XPS 13 tonight. However, the advent of skylake has me considering waiting (but who knows if that'll be out in the next 6 months), as does your comment.
The rumour is that the next-gen macbook air will have it, and that this is going to happen this quarter. Iff that happens, then other manufacturers are going to hop on board too.
If not, I expect it may take over a year before it's ubiquitous -- Intel isn't including USB 3.1 in Skylake, so it will require an extra chip, something manufacturers are very hesitant to add to laptops.
Hmm. Rumor has it the next Air refresh is next week. If that's not the next-gen one, then I'm a bit skeptical that they'll put a brand new model out right after a minor change. I guess I'll see what happens next week and then make my decision.
I have another OS installed in a VM, an admittedly ridonkulous number of open tabs (that's just how I work) and an IDE, and that puts me at around 11GB in use.
I had the same impression, 8G max is a regrettable decision. The price is also a little too much and Dell seems to be one of those coupon bullshit companies.
