Become a MacRumors Supporter for $50/year with no ads, ability to filter front page stories, and private forums.
Apple releases their iOS development tools only for OSX, and iTunes only for OSX and Windows.

First you say you'd have left Mac a long time ago if Linux had more software, then you turn around and use that reasoning I just quoted to (apparently) justify staying with the Mac. When you've made up your mind once and for all, let someone know.
 
First you say you'd have left Mac a long time ago if Linux had more software, then you turn around and use that reasoning I just quoted to (apparently) justify staying with the Mac. When you've made up your mind once and for all, let someone know.

Isn't what I mentioned software?
 
considering PCs are still hanging out in the 8gb RAM territory, I find 32gb RAM something that won't happen for a while.
 
32GB is fundamental for bringing back a workstation status to the Macbook Pro.
I dont know what you are talking about.
I only have the mid 13" config with 8gb and this is by far the most potent laptop i ever had.
It can multitask video and photo editing without breaking a sweat.
I have personally never seen a laptop that can do this.
 
I dont know what you are talking about.
I only have the mid 13" config with 8gb and this is by far the most potent laptop i ever had.
It can multitask video and photo editing without breaking a sweat.
I have personally never seen a laptop that can do this.

What if you need dealing with 20GB of data without swapping? This happens when you need to retrieve millions of rows in a database to convert to another format or you need multiplying very big matrices. Ok, these are not common scenarios, but they happen rather frequently at scientific research.

Maybe Apple should produce a Macbook called "Macbook Sci". It would come with discrete GPUs (preferably nVidias so we could opt between OpenCL and Cuda), 16GB of RAM (with BTO options of 32GB and 64GB), a 17" display and a hexacore i7. 3-hours of battery life with all these components turned on or up to 7-hours when doing just browsing/reading PDFs/writing.

----------

Agreed. Would be great for a 17" I think. But that's just me wishing the 17" would come back. :D

Yes. That would be nice... with a thinner bezel it would look practically the same as the cMBP 15".
 
What if you need dealing with 20GB of data without swapping? This happens when you need to retrieve millions of rows in a database to convert to another format or you need multiplying very big matrices. Ok, these are not common scenarios, but they happen rather frequently at scientific research.

Stop using suboptimal algorithms. Neither large data aggregation nor big matrix multiplication requires you to have all the data in the RAM all the time. If you are using R (which you should), look at the packages bigmemory and ff packages, as well as their friends.
 
I dont know what you are talking about.
I only have the mid 13" config with 8gb and this is by far the most potent laptop i ever had.
It can multitask video and photo editing without breaking a sweat.
I have personally never seen a laptop that can do this.

Just because your machine "feels fast" to you, doesn't mean that it's acceptable for professional users. A dual core i5 and 8GB of RAM is, if not less than the bare minimum, the bare minimum for editing video. I use After Effects professionally, and 8GB would not really be in the acceptable range for what the program demands. Uncompressed HD or 4K frames take up a lot of space, and I can fill up all 96GB of my work machine's RAM, previewing less than a minute of animation.
 
Just because your machine "feels fast" to you, doesn't mean that it's acceptable for professional users. A dual core i5 and 8GB of RAM is, if not less than the bare minimum, the bare minimum for editing video. I use After Effects professionally, and 8GB would not really be in the acceptable range for what the program demands. Uncompressed HD or 4K frames take up a lot of space, and I can fill up all 96GB of my work machine's RAM, previewing less than a minute of animation.
No doubt. But laptops have never been able to carry that sort of workload.
 
Stop using suboptimal algorithms. Neither large data aggregation nor big matrix multiplication requires you to have all the data in the RAM all the time. If you are using R (which you should), look at the packages bigmemory and ff packages, as well as their friends.

I'll take a look. Anyway I can't imagine how a swapping-based approach for dealing with large datasets can make things significantly better (that is, way faster), but thanks for the advice, I'll try it.
 
I'll take a look. Anyway I can't imagine how a swapping-based approach for dealing with large datasets can make things significantly better (that is, way faster), but thanks for the advice, I'll try it.

Not sure what you mean by 'swapping-based approach', but the techniques I have referenced don't necessarily make things faster. They simply allow you to work with big data without having to load the full dataset into RAM; in another words, only the data you need right now is loaded. On a machine with tons of RAM, there shouldn't be any speed difference. On a machine with little RAM, there will be a dramatic improvement.
 
I would have thought it was obvious that different Pro users have different requirements, Apple market a range of options that make the MBPs suitable for a range of Pro users but have never claimed to cater for every single Pro usage out there, that is simply a diminishing-return market (as you head towards more/bigger/faster), for which Apple probably assesses the prime equipment being a desktop config rather than a laptop.
 
I would have thought it was obvious that different Pro users have different requirements, Apple market a range of options that make the MBPs suitable for a range of Pro users but have never claimed to cater for every single Pro usage out there, that is simply a diminishing-return market (as you head towards more/bigger/faster), for which Apple probably assesses the prime equipment being a desktop config rather than a laptop.

Apple caters to even less people with the rMBP than with the cMBP.
 
Stop using suboptimal algorithms. Neither large data aggregation nor big matrix multiplication requires you to have all the data in the RAM all the time. If you are using R (which you should), look at the packages bigmemory and ff packages, as well as their friends.

What he says. Back in college decades ago I worked in image processing on a 64k minicomputer with images between 64-128k. With OS and programs, you had about 40k left. You had to think about how you were processing your data.

Nothing like working with sub-optimal systems to help you work more efficiently.
 
What he says. Back in college decades ago I worked in image processing on a 64k minicomputer with images between 64-128k. With OS and programs, you had about 40k left. You had to think about how you were processing your data.

Nothing like working with sub-optimal systems to help you work more efficiently.

Well, I programmed in x86 assembly and simulators of earlier architectures. I don't need moral lessons from Apple fanboys. If you want to live handling as much data and as much fast as Apple allows, good for you!
 
I have been needing massive storage and 32gb ram as well and I have been using a Thinkpad W530 for just that purpose. I require multiple virtual machines and heavy photoshop work and both of those need massive ram. Yes, ideally this is done better on my Mac Pro 12 core but when you gotta go mobile the thinkpad takes the cake. I won't be buying another thinkpad though as the W540 is reported to be a massive dud so I will hold out with this one I have until apple releases a 32gb ram capable Macbook pro then I can dump this one and give it to the my kids.
 
Apple caters to even less people with the rMBP than with the cMBP.
No. The haswell pros are light, with an excellent display and ultra long battery life.
They handle medium load multimedia work and "normal" stuff excellently.
The 2012 mbp didnt support 32gb ram either and to get more than 2tb of ssd storage you would need more sata slots.

Laptops never have been desktop replacements. :apple: hasnt changed anything in this regard.
People who need tons of ram and huge amounts of storage on a laptop are a tiny, tiny niche market.

What do you guys even do? Cgi for blockbusters? On a notebook?
 
Last edited:
No. The haswell pros are light, with an excellent display and ultra long battery life.
They handle medium load multimedia work and "normal" stuff excellently.
The 2012 mbp didnt support 32gb ram either and to get more than 2tb of ssd storage you would need more sata slots.

Laptops never have been desktop replacements. :apple: hasnt changed anything in this regard.
People who need tons of ram and huge amounts of storage on a laptop are a tiny, tiny niche market.

What do you guys even do? Cgi for blockbusters? On a notebook?

Got to agree with this. If you really, absolutely must have 32GB+ then you really ought to be using a desktop. 32GB will come, potentially with Broadwell or Sylake an an option, but it is unnecessary to the point of stupid for the vast majority of their pro users. I do appreciate some people might have uses for 32GB or more of memory but like I said, who really needs that? I'm sure 16GB will suffice, even for those that may potentially make use of more, plus you have a monstrously fast SSD so streaming data from there isn't as painful a pill to swallow like in the good old HDD days. The bigger issue I'm surprised your not talking about is that the RAM in a rMBP only runs at 1600MHz which is a bit weak. From what I've seen 1866MHz is the sweet spot.

Come to think of it I don't think I've even seen a 16GB DDR3 DRAM stick, do Micron etc build chips big enough to do that yet?

With systems heading to a unified memory system between CPU and GPU high bandwidth memory is going to become more important than volume in my opinion.
 
Register on MacRumors! This sidebar will go away, and you'll see fewer ads.