Celeron 950 rated as barely PIII500, when it shoul

Alexis

Distinguished
Oct 28, 2001
157
0
18,680
Hi all,
After reading THG article about Celeron overclocking I was very pleased to discover that my old BH6 Abit board (BX440 chipset) was able to overclock Celeron633 up to 950MHz after a BIOS update. So did I. And proceeded to benchmark the new system in Norton Utilities. My previous Celeron 450’s result was 145 points while the new one “climbed” only to 215 points, equalling in power a PIII500. I thought it would go at least to PIII700.
In another THG article I saw the 133 BX chipset even faster than the i815. Well, my config is based on SDRAM 100 (256Mo) but still I can’t explain those results. Does anybody can?
 

NickM

Distinguished
Mar 25, 2001
563
0
18,980
Sorry, I’m not the right guy to lit on synthetic benchmarks, I have the NortonSystemWorks but I’ve never taken seriously what their System Information Benchmark said to me. As for the <A HREF="http://www.sisoftware.demon.co.uk/sandra/" target="_new"> SiSoft Sandra </A>, it wasn’t able to recognize the right speed for my AMD, showing and benchmarking it at least twice higher.
By the way, while more competent people are preparing the answer, could you provide what the Sandra shows for you? Probably, you’ll find a difference.
I prefer to do my comparisons based on some real software I work with.

And one more thing… My opinion on Celerons is different, unless it is a Tualatin.
Believe or not, my Pentium II o/c 333 MHz @ 133FSB is at least twice faster than any Celeron 466 –500 with the same clean install of Oracle we have in school. We checked, and our SysAdmin confirmed that.
 

Alexis

Distinguished
Oct 28, 2001
157
0
18,680
"Believe or not, my Pentium II o/c 333 MHz @ 133FSB is at least twice faster than any Celeron 466 –500"
I agree. My PII 300 is only 10 points under Celeron 450 (135/145).
 

Alexis

Distinguished
Oct 28, 2001
157
0
18,680
I can see only three issues here.
1) Newer chipset architecture
2) Higher memory speed (133 over 100)
3) Different Motherboard implementations.
But again does it explain THAT difference in results?