![]() |
|
|
#1 |
|
Feb 2003
2·59 Posts |
I have a MSI KT3 board with this chipset. I got a SUM(INPUTS)!=SUM(OUTPUTS) error and did some digging work. The system seemed very stable, passed all the torture tests (even the large inplace FFT's) and passed memtest86. Still ... this error buged me. Now I think I found what it was. This chipset allows you to use 333MHz memory with a 266MHz FSB CPU (I have a Athlon 2400+). Everything works fine if you let the MB to choose the CAS timings. For the PC2700 Samsung memory that I have the MB chosses CAS2.5 . I wasn't happy about it and changed it to CAS2 some time ago.
There is a french hardware-oriented site that did extensive testing on ABIT and MSI motherboards that use the KT333 chipset. They say it won't work 100% stable with PC2700 memory and CAS2 no matter what. CAS2 will work good with PC2100 only, for PC2700 let it be 2.5 like the MB wants. |
|
|
|
|
|
#2 |
|
Oct 2002
Lost in the hills of Iowa
26·7 Posts |
It depends on the memory - my ASUS A7V333-X runs CAS2 with a Thoroughbred XP1800+, but I think I'm running either high-end PC2700 CAS2-specific DDR, or PC3000.
MIght also be that ASUS is better designed on the memory traces, but since ABIT has at least as good a rep in that department, I doubt that's the problem. Also, VIA was supplying some of the last KT333 chipset orders with re-labled KT400 chipsets, rather than keeping all of their KT333 chipset manufacturing going to support a dying chipset. IIRC the KT400 was a lot better at handling a mismatch between memory and CPU clock speeds. |
|
|
|
|
|
#3 |
|
Sep 2002
Austin, TX
10001100012 Posts |
I would choose the higher clock speed over the lower CAS lat. Too bad most KT333 boards don't take DDR400 at full speed
.
|
|
|
|
![]() |
Similar Threads
|
||||
| Thread | Thread Starter | Forum | Replies | Last Post |
| Celeron D and Willamette owners | Prime95 | Software | 10 | 2005-07-08 08:44 |
| e-mail to supercomputer owners | ixfd64 | Lounge | 17 | 2004-11-21 20:22 |