Coding and big data on 2016 nTB

[gerlof92]

Hi everyone,

Right now I am using a 2011 15" MacBook Pro with 16gb ram and SSD.

I am thinking of upgrading towards the new 2016 nTB MBP. I mainly want a smaller and more portable MacBook.

My only concern is wether the 2016 nTB MacBook is capable of doing big data analytics with R, Weka and Python or other apps which are associated by this. Can anyone of you with data experience tell me if programs like these will run as smooth on the 2016 2.0 ghz i5 in comparison with a quad core for example.

Will I notice a difference either positively or negatively?

Thanks,

Gerlof

 

[RubenNick]

Ram is the main concern in big data analysis, if you are currently getting by with 16gb the new one with 16gb will handle just fine.

In terms of CPU differences, it highly depends on what you are running. R and Python are languages so I suppose if you are running codes written in those languages then how parallelizable the codes are i.e. if your code can only be split into two processes actively being performed, then having quadcores (or even 128-cores) will have no effect on efficiency.

Can't say much about Weka as I have never used it. But if from what I read is true, then it does utilize quad-core cpu. So to answer your question would you notice any difference switching from a 2011 quadcore CPU to a 2016 dual-core CPU? Most likely, as quad-core would likely out-perform dual-core since intel CPU clock speed/cpi haven't improved that much the past couple years. So yes it'd probably negatively affect the execution time.

I wouldn't switch to a new MBP if I were you since your current laptop seems to be running the programs smoothly as is. In all honesty you won't see much of a difference in speed even if you switch to the new 15inch ones (RAM and CPU improvements are not that substantial if at all). I'd wait till Apple release at least a 32gb ram option to consider getting a new one (well personally I'd want them to give me the retina form factor with improved CPU, GPU, and RAM).

 

[gerlof92]

Thanks for your reply!

 

[leman]

I was running data mining on the full google corpus on a 8gb MBP using R... the question is really how big your data is and how do you work with your data. If your algorithms need to load everything at once, you'll need a supercomputer. But if you can work with the data one slice at a time, the MBP is a wonderful machine.

P.S. the largest data set that I work with in R that is comptrjekyvresudebt in memory is around 12GB uncompressed. This is in a 2015 16Gb 15" model. Works well enough.

 

[gerlof92]

P.S. the largest data set that I work with in R that is comptrjekyvresudebt in memory is around 12GB uncompressed. This is in a 2015 16Gb 15" model. Works well enough.

So a 16gb will be definitely the way to go if I am considering a new Macbook?

 

[leman]

So a 16gb will be definitely the way to go if I am considering a new Macbook?

If you buy a new 15" (and I would def recommend you to get a quad core for these tasks), 16GB is the only option anyway.

For heavy duty tasks you'd still want to use a supercomputer. If you don't have access to one, much can be done using Amazon cloud computing.