1

I do a lot of statistical computing in R, particularly text analysis which involves a lot of sparse matrix operations and EM algorithm calls. What hardware specs are most important for optimizing these procedures? Processing power, memory, cores?

Improve this question
2
  • This also depends on how well the algorithms you have make use of multi-core CPUs (c.f. Amdahl's law)
    – slhck
    May 29, 2012 at 20:29
  • @slhck Yes, I should say that most of these are not multi-core optimized as of yet but it is moving in that direction.
    – Trey
    May 29, 2012 at 21:32

1 Answer 1

Reset to default
1

If the data processing is mostly calculations on a small or moderate amount of data your processor would be the most important. As the amount of data you are using raises, the importance of disk access speed and RAM grows.

Improve this answer
1
  • I would add cache size.
    – Yitzchak
    May 29, 2012 at 21:32

Not the answer you're looking for? Browse other questions tagged .