Random Phrase: Home sweet home...


Read Me First!

This is the Spode's Abode archive. The old articles and forum have been kept here purely for historical purposes and are no longer updated.

Not all portions of this archive may work as expected.


Please visit the new site.



TMC/DFI AM75-EC (KM133)
Written by Spode (17/Mar/01)
Page 2 of 4

Untitled Document
The KM133 chipset (VT8365) is basically the KT133 chipset with onboard Savage 4 (or ProSavage as they now call it). But on the via site, it mentions "Flexible 66/100/133MHz FSB settings" - yet we all know the KT133 only supports 100mhz officially and only up to around 112 unoffically. Seeing as there was no multiplier adjustment on this board, overclocking FSB was my only option - which luckily this board supports!

I was expecting this to overclock about the same as a KT133 chipset like my old Abit KT7 and Aopen AK73 Pro. But on the otherhand, the VIA site said it supported 133mhz FSB (but no mention of support for 133mhz chips).

The board supports a maximum of 166mhz FSB in 1mhz increments, but I wasn't expecting to go that high at all. But It was quite happy to run at 133mhz on my Duron 700. This gives an overclock of 933mhz. For chips with low multipliers, an extra 33mhz FSB is not enough to max out your chip, but chips with higher multipliers seem a little closer.

Multiplier
100mhz
133mhz
6x
600mhz
800mhz
6.5x
650mhz
866mhz
7x
700mhz
933mhz
7.5x
750mhz
1000mhz
8x
800mhz
1066mhz
8.5x
850mhz
1133mhz
9x
900mhz
1200mhz

As you can see, with the higher multipler chips - overclocking in this board is quite acceptable. (just before I fried it, my Duron 800 was running at 129mhz happy). As far as Tbird overclocking - this board only has a 2 phase power solution, so I'm not sure how happy it would be with the faster chips (as they might not get enough current).

There was no voltage modification on this board, but the great thing about AMD chips, is the ability to just connect together the L7 bridges with a pencil and get us 1.85volts (read 1.92v in the bios). Of course, with some of these newer Tbirds coming in that need a lower voltage to overclock (as discussed a lot in the overclockers.com forum) - this may be a problem which will involve some cutting with a scalpal.

Now, 133mhz from something based on the KT133 chipset seems a little high, yes? Sisoft Sandra reported it to have the KT133 chipset, and also said my PCI was running at 45mhz. WCPUID reported it to have the KT133A chipset. Either way, the KM133 chipset seems to be a relatively good overclocker. I couldn't get it stable at anything higher than 135mhz and posted no higher than 140mhz.

I think the reasoning for not being able to go any higher is PCI speed. Sisoft said I had a PCI speed of 45mhz. So when running my memory at 3x PCI, I should get 133mhz? Which I did. Which shows the main problem with this chipset - NO 1/4 PCI DIVIDER. Of course, wether that is a limitation of the board, or the chipset - I don't know. What I do know, is that there will also be a KM133A chipset, which will of course have the 1/4 PCI.

Here are some benchmarks of the Duron 700 running under Windows 2000 with 256mb of ram and an ATI Radeon 32mb DDR.

Sisoft CPU (ALU/FPU)
Sisoft Memory (ALU/FPU)
Quake 3 Fastest
7x100, mem: PCIx4
1962/958
387/410
83fps
7x133, mem: PCIx3
2610/1276
402/411
98.8fps

Not a bad little overclock...


<< Back | 1 | | 2 | | 3 | | 4 | Next >>



Copyright Andrew Miller
Please read our disclaimer

Search the site:

Random

MIST Silent Twin Cooler