Super User is a question and answer site for computer enthusiasts and power users. Join them; it only takes a minute:

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

My CPU model is 2 and a half years old now, released Intel Q4'2010. How can I determine if getting a new CPU will improve how fast I surf the Internet, versus getting a better ISP subscription?

share|improve this question
up vote 3 down vote accepted

Measure what your current bottleneck is. Run things like vmstat (BSD), iotop and top (Linux), taskmanager or resource manager (windows). Then start your browser and watch what the measurements do while you surf.

If the CPU tends to spike a lot to 100%, or of it stays at 100% for periods, then a faster CPU will help. If it never even spikes above half load then a faster CPU will not help much.

Do the same with memory: Does opening a few dozen tabs cause you to page (swap)? In that case more memory might help (adjust for your normal usage: Mine is between 30 and 60 open tabs at the same time).

If neither of the above two is stressed then the network might be the weakest link. However that can have several reasons. Pure throughput is just one of them.

Other reasons could be access times, or no installed adblockers (many webpages wait for advertisement to load. And the servers providing the adds are often damn slow. Which means that actualy surfing can feel a lot faster with an add-blocker. Even if your thoughput is more than enough to handle the normal surfing data and the adds at the same time).

There are probably a few more points which I have forgotten, but the main thing I tried to say is "measure!". (Or as we Dutch say 'Meten is weten' ('to measure is to know').

Tl;dr: A faster CPU is probably not going to help much. But check how much CPU is used with your current setup.

share|improve this answer

You must log in to answer this question.

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