Statistics and the Core2 vs. i5 argument-which one.

[GKC]

OK, here's the question for those of you who may be doing something similar: I do a lot of Stats (R mostly). I have the quad core linux desktop to run calculations overnight. However most of my actual work is done at Univ away from home. I have loved the unibody 13" (4meg)(before it graduated to pro) for the portability. Getting a new something or other anyway-debating if I should jump to the next or live with bogging down.

The benchmarks however are highly centered around doing work with graphics/gaming/video etc. The calculations I have it doing involve a blinking cursor and listening to the fans roar.

Will there be a noticeable difference with pure number crunching to justify the loss in portability and extra money or is it something along the line of waiting 16min for the MCMC chains rather than 17?
Thanks.

 

[pjcforpres2020]

It depends on if the programs even take advantage off multi core processors. I would think no considering mathematics is very linear, and while they could design the program to do an equation per core, I doubt they have.

 

[TrojanX]

Well the first thing to consider is what the I5 really is. Virtually it is a 2.x Ghz C2D processor that has been "Virtually" split into 4 cores using a process called hyperthreading. I don't know how your number crunching (Excel ect) but I can almost promise you that you will barely benefit (at least for number crunching) from the I5. The reason I say this is that the hyperthreading only loads the secondary cores with mostly visual things such as photos/videos and unless the programs are telling the computer to send the numbers through the virtually separated cores than it will revert to the C2D. Go with the 13" 2.4 honestly. Less bugs than the I5s and almost the same power for your uses.

 

[elleana]

What program do you use to do the number crunching?

 

[Construct]

What program do you use to do the number crunching?

He said he uses R in the original post.

 

[GKC]

Thanks for the replies.

I did look this up, seems like there's some multicore going on with matrix calculations.

The MacOS port of R on CRAN is linked to ATLAS, a "tuned" BLAS that uses multiple cores for computations. As a result, R on a multi-core Mac (as all new Macs are these days) really zooms. But the Windows binaries on CRAN are not linked to an optimized BLAS. It's possible to compile and link R yourself, but it can be tricky.

On the other hand using the money to get the 13" up to 8 gigs or even a SSD might have more bang in usability and multitasking with some windows programs I often have to use in parallels. .

Going to lug around a friends 15" (yes a MBP, not something else) and see the difference in portability first hand.

Thanks.