SPONSORED LINKS
I just timed my Thinkpad 760CD (120MHz, 40MB) for a build and timing test of a project, and my Zeos 100MHz Pentium 32MB home system on the same. In both cases, I rean the make twice, so all pertinent headers etc. would be in the caches. This means disk spead wouldn't be a factor. Results: 2 min 20 second for the tower, 4 min 20 second for the laptop. The test suite put out alot of text which scrolled on 80x40 xterms, so graphics performance was involved. A part of the test that timed itself showed the laptop actually performing 10% faster on a read-only section of the test, but 10% slower during write- intensive parts. (insert and delete times for a balanced binary tree were faster on the tower, and the searches which don't write memory were faster on the laptop.) - - - - - What is the point? Well, if you are thinking about buying a laptop vs. a PC, you should know that the laptop architecture cuts corners on performance. How? I suppose, either no L2 cache, no EDO ram, or both. There might be other problems with the buses or something. Also, I did timing tests on an NEC Versa 100MHz 486 with 24MB, that were not much slower than my laptop.. If/when I buy my own laptop, I will probably aim for lower performance processor, if it is any cheaper or (more to the point) faster delivery. The memory and hard disk speeds and graphics are the bottlenecks on these things, most of the time... Frank