Mathematica benchmark on two different system

benchmarkinghpcintel-core-i7windows 7xeon

I have two system one has a Xeon E5-1650 CPU, 24GB of RAM, and a 7200rpm hard drive. The other system which is a Lenovo G510 laptop, has a core i7-4700MQ CPU, 4GB of RAM, and 5400rmp hard drive.

The result of the benchmark for Xeon system:
enter image description here

The result of the benchmark for i7-4700MQ system:
enter image description here

It seems in spite of the fact that Xeon has much higher clock speed, as much as twice, the Mathematica software performs better on a system with much lower RAM and supposedly weaker CPU!
Could somebody explain it to me why the result of the benchmark is as it is? I am wondering if I could buy a cheaper laptop and still Mathematica could have an acceptable performance. It seems stronger CPU does not always lead to a faster computation. If I knew the important factor, I could buy a more suitable laptop for myself.

Here are the timing for each test and a description of the test:
enter image description here

Best Answer

The reason for this is that a computer is only as fast as its slowest component!

One of your machines may have a much faster CPU, a shed load more RAM and a faster bus speed, but all of that is nearly worthless if the hard drive only spins at 5400rpm and can only move data from the drive to the processor at a slow speed!

Mathmatica is quite a drive/swap intensive program - it needs a fast hard drive. If you put a 15k drive - or better yet, an SSD instead of a 5.4k in the faster machine, it would absolutely fly!

Its the same as having a forumular 1 racing car, but sticking it on a 30mph road - you still have the fastest theorewtical car - but its worthless if the road is limited! (weak metaphor, but I haven't had a coffee in a while)

Related Question