Re: Perfomance problem backupscript (more details and tests)
[email protected] Tue, 18 Oct 2005 18:43:04 +0200
| Newsgroups | gmane.comp.archivers.star.user |
|---|---|
| Message-ID | <[email protected]> |
Hi, > > > In 1993 [...] Linux was completely unuseable > > > At that time, Pentium CPUs have been _boringly_ slow.... > > > > Actually already 80486 came quite near to the performance > > of the SPARC CPUs of that time. Pentium then caught up > > with performance-per-MHz and pulled away with the MHz. > > Benchmarks for Intel systems at that time did not match > real world performance tests. I know those bonnie et.al. results of iX Magazin in the early 90s where a 386 allegedly outperformed a SPARC CPU of the same clock speed. That's not what i am talking about. I talk about the Unix workstations i made my living from in 1993 and 1994. I talk about real work benchmarks with real work programs and real work users. An unrealistic benchmark would have been recognized by those users. At that time PCs for 8000 DM could compete with the SUN shoebox workstations which were about 25000 DM. The pizzaboxes tried to play in the SGI class (50000+ DM) but could not compete there. Both, x86 and Linux made a surprising development. I am not bound to any CPU architecture. I never was. But since ten years whenever i considered to buy a new computer i ended up with Intel or AMD : Pentium/133 , AMD/350 , Pentium/669 , AMD/1900 My last personal workstation before PC was a SUN IPX/40. I considered MIPS, Alpha and PPC (each time there was another best processor on the scene). None was able to provide a useable computer system at a competitive price-performance ratio. If SUN or SGI had sold their own CPUs within systems that where not blatantly overpriced, then Intel would never have been able to extend its unreasonable architecture to such a tremendous performance. SPARC and MIPS were of course much better than x86. They could be produced cheaper and delivered more performance per MHz. Nevertheless Intel allowed anybody to build cheap systems around x86 and showed a ridiculous effort to make its own clumsy chips faster. The self exploitation of the PC manufacturers helped Intel to make a business out of CPUs which costed a multiple of their competitors without being better. But without being much worse either. The negative effect is that there is not much fun in being CPU independend, anymore. :( The positive effect is that my most recent computer did cost less than the keyboard of my first professional workstation. Have a nice day :) Thomas