sis540 chipset
sis540 chipset
hehe.
got an sis540 board from china yesterday. its a m581lmr from pcchips, mATX with 2 pci and one adimm slot. my motivation was just testing it for higher fsb frequencies, the winbond w83194r-630a pll supports also 112/37, 133/33 140/35 and 166/33 mhz cpu/pci settings (16 in total).
ok. first obervation @ 100mhz - very poor memory performance. i will post some benchmarks later, it was more than disappointing.
the ami bios on that board was dated 06/26/2000. it does not allow settings over 100mhz and there are no jumper settings on the motherboard. what i did is flash an award bios v6.00pg for the winfast 5400max on it (also sis540) - and it works. lol, it happend faster i wished, wanted to save the old bios but unfortunately the ami flash prog has simply overwritten it.
the silly thing is - the winfast board has a cypress c9630 pll. it has 32 possibilies for cpu/sdram/pci clock setting. that award bios alows to set them by software, the multiplyer only by hardware (jumpers).
thats why my board starts with 2,5x everytime with every fsb i wish.
some explanation here - the winbond pll has got 4 bits of fsb selection in the cpu frequency select register. the cypress got 5 bits, the first four on same position as for the winbond and nr. 5 on the spread spectrum modulation position. as it is quite similar, it luckily works ...
the fastest i got was 350 mhz (2,5x140), but couldnt get into windows, 166 fsb booted, but with boot block errors. maybe its the ram. no idea yet.
133 fsb works with errors only in win98. maybe there is some trick to get it working stable, i know that sis530 worked well. any suggestions?
@uranium - i know u are using the powecolor mainboard. i think it would be no prob flashing the ami bios on it. which pll is used on the p561a?
got an sis540 board from china yesterday. its a m581lmr from pcchips, mATX with 2 pci and one adimm slot. my motivation was just testing it for higher fsb frequencies, the winbond w83194r-630a pll supports also 112/37, 133/33 140/35 and 166/33 mhz cpu/pci settings (16 in total).
ok. first obervation @ 100mhz - very poor memory performance. i will post some benchmarks later, it was more than disappointing.
the ami bios on that board was dated 06/26/2000. it does not allow settings over 100mhz and there are no jumper settings on the motherboard. what i did is flash an award bios v6.00pg for the winfast 5400max on it (also sis540) - and it works. lol, it happend faster i wished, wanted to save the old bios but unfortunately the ami flash prog has simply overwritten it.
the silly thing is - the winfast board has a cypress c9630 pll. it has 32 possibilies for cpu/sdram/pci clock setting. that award bios alows to set them by software, the multiplyer only by hardware (jumpers).
thats why my board starts with 2,5x everytime with every fsb i wish.
some explanation here - the winbond pll has got 4 bits of fsb selection in the cpu frequency select register. the cypress got 5 bits, the first four on same position as for the winbond and nr. 5 on the spread spectrum modulation position. as it is quite similar, it luckily works ...
the fastest i got was 350 mhz (2,5x140), but couldnt get into windows, 166 fsb booted, but with boot block errors. maybe its the ram. no idea yet.
133 fsb works with errors only in win98. maybe there is some trick to get it working stable, i know that sis530 worked well. any suggestions?
@uranium - i know u are using the powecolor mainboard. i think it would be no prob flashing the ami bios on it. which pll is used on the p561a?
- Uranium235
- Senior K6'er
- Posts: 183
- Joined: Mon Sep 12, 2005 9:59 pm
no results yet,
but one thing to clearify here about sis540 mATX boards from pcchips and matsonic.
from each manufacturer there are two different versions availible. they look very similar, matsonic boards seem to be rebranded versions of pccchips ones. they are availible with lan onboard and without:
pchcips - m581mr and m581lmr ---- (Lan)ModemRiser
these boards are availible from 512k to 2m of cache.
matsonic - ms6393d and m6393e - only e with lan onboard
these have only 512k of cache!
not only the cache is different. some capacitors near the cpu are also not on same place. but it looks damn equal!
pcchips m581mr (wrong motherboard description!!!):
http://www.yjfy.com/M/Matsonic/mainboard/MS6393D_E.htm
matsonic ms6393e:
http://www.motherboard.cz/mb/matsonic/MS6393E.htm
hard to see, but they do differ. thats why there are a lot of sites confusing these two.
but one thing to clearify here about sis540 mATX boards from pcchips and matsonic.
from each manufacturer there are two different versions availible. they look very similar, matsonic boards seem to be rebranded versions of pccchips ones. they are availible with lan onboard and without:
pchcips - m581mr and m581lmr ---- (Lan)ModemRiser
these boards are availible from 512k to 2m of cache.
matsonic - ms6393d and m6393e - only e with lan onboard
these have only 512k of cache!
not only the cache is different. some capacitors near the cpu are also not on same place. but it looks damn equal!
pcchips m581mr (wrong motherboard description!!!):
http://www.yjfy.com/M/Matsonic/mainboard/MS6393D_E.htm
matsonic ms6393e:
http://www.motherboard.cz/mb/matsonic/MS6393E.htm
hard to see, but they do differ. thats why there are a lot of sites confusing these two.
Hardly surprizing since most people have never seen any of them.
Superpuppy 3
K6-3+ 450 ACZ (6x100)
DFI K6BV3+/66 Rev B2 (2 Meg) w/ 2x28mm Chipset Fans
2x256 Meg PC 133 Hynix SDRAM
1x 20G Maxtor (7200)
2x 80G Maxtor (7200) Ducted w/ 2x486 Fans Mount
52/24/52/16 LG CDR/RW/DVD
8/4/3/12/24/16/32 LG Super Multi
ATI 9000 aiw Radeon AGP
SB Audigy 1 MP3 Sound
CMD 649 IDE Controller
NEC USB 2 Card
K6-3+ 450 ACZ (6x100)
DFI K6BV3+/66 Rev B2 (2 Meg) w/ 2x28mm Chipset Fans
2x256 Meg PC 133 Hynix SDRAM
1x 20G Maxtor (7200)
2x 80G Maxtor (7200) Ducted w/ 2x486 Fans Mount
52/24/52/16 LG CDR/RW/DVD
8/4/3/12/24/16/32 LG Super Multi
ATI 9000 aiw Radeon AGP
SB Audigy 1 MP3 Sound
CMD 649 IDE Controller
NEC USB 2 Card
so, results for sis540.
... first statement: this chipset is horribly slow. at least with the default settings in bios. with a k6 2+ @100x6 the performance is not better than with a normal cyrix cpu at 250mhz on a mvp3 system. here are some benchmark results:
motherboard: pcchips m581mr, 512kb l3 cache
cpu: amd k6 2+ @ 100x6
ram: 2x256mb apacer cl2 @ 3-3-3-6
graphic: sis540 graphic unit with 64mb shared ram.
win98se
everest
read 126
write 78
delay 542,8
superPi 1M
9:36
cpumark99
25,1
as you clearly see, disgusting. in the original bios (ami) there are nearly no settings to improve those resulst, at least for this board. max fsb setting is 100mhz, and only can be switched in bios (no hardware setting!)
thats the negative. now the positive, why it made me so interested in that board/chipset.
1. the pll (windond 630a) is able to provide asynchronous cpu/sdram frequencies and is 166mhz capable (!!!).
2. agp 4x is realy enabled - not like for the alladin7. so it has a graphic unit based on the 64bit sdram architecture and is capable to run even with 150mhz sdram clock.
with some ideas i started to test these capabilites and first flashed a new bios. what i can advise everbody out there is the winfast 5400max bios, award 6.0; much more settings for sdram timings, complete fsb switching up to 166mhz (with the /5 divider). now the fun should begin ....
... and it did not. i tried the best i could, but the system is only stable till 112/112 host/sdram, even with cache disabled. if the cpu host frequency is higher, results are no boot or crash in windows or in loading it. only at 166/166 the post without logo or selftest appeared:
hard cmos checksum error.
please insert disk in drive a:
loading dos = freeze.
for asynch mode it becomes better. the max stable frequency for sdram i reached is with the 100/150 setting. the best overall results are:
everest
read 167
write 112
delay 386,2
superPi 1M
7:43
cpumark99
29,1
note that even with onboard graphic and cache enabled the system is stable @150 mhz sdram. as well for 97/129 and 83/125.
i have to admit, it was more playing around than a systematic testing of what every setting gives. the number of factors is huge, and i did not found the crucial option to increase system stability in a noteable manner. even using a pci video card gave similar results, allthough with bit higher benchmark scores.
no idea, if somebody will try to do same work, but me, i will finish testing the sis540 for now. im defeated ...
... first statement: this chipset is horribly slow. at least with the default settings in bios. with a k6 2+ @100x6 the performance is not better than with a normal cyrix cpu at 250mhz on a mvp3 system. here are some benchmark results:
motherboard: pcchips m581mr, 512kb l3 cache
cpu: amd k6 2+ @ 100x6
ram: 2x256mb apacer cl2 @ 3-3-3-6
graphic: sis540 graphic unit with 64mb shared ram.
win98se
everest
read 126
write 78
delay 542,8
superPi 1M
9:36
cpumark99
25,1
as you clearly see, disgusting. in the original bios (ami) there are nearly no settings to improve those resulst, at least for this board. max fsb setting is 100mhz, and only can be switched in bios (no hardware setting!)
thats the negative. now the positive, why it made me so interested in that board/chipset.
1. the pll (windond 630a) is able to provide asynchronous cpu/sdram frequencies and is 166mhz capable (!!!).
2. agp 4x is realy enabled - not like for the alladin7. so it has a graphic unit based on the 64bit sdram architecture and is capable to run even with 150mhz sdram clock.
with some ideas i started to test these capabilites and first flashed a new bios. what i can advise everbody out there is the winfast 5400max bios, award 6.0; much more settings for sdram timings, complete fsb switching up to 166mhz (with the /5 divider). now the fun should begin ....
... and it did not. i tried the best i could, but the system is only stable till 112/112 host/sdram, even with cache disabled. if the cpu host frequency is higher, results are no boot or crash in windows or in loading it. only at 166/166 the post without logo or selftest appeared:
hard cmos checksum error.
please insert disk in drive a:
loading dos = freeze.
for asynch mode it becomes better. the max stable frequency for sdram i reached is with the 100/150 setting. the best overall results are:
everest
read 167
write 112
delay 386,2
superPi 1M
7:43
cpumark99
29,1
note that even with onboard graphic and cache enabled the system is stable @150 mhz sdram. as well for 97/129 and 83/125.
i have to admit, it was more playing around than a systematic testing of what every setting gives. the number of factors is huge, and i did not found the crucial option to increase system stability in a noteable manner. even using a pci video card gave similar results, allthough with bit higher benchmark scores.
no idea, if somebody will try to do same work, but me, i will finish testing the sis540 for now. im defeated ...
The things that immediately occur to me are : What voltage were you running the processor at while you were attempting to reach these FSB speeds ? And what quality of SDRAM were you using ? Did you try using WPCredit to experiment with any of the chipset settings ?
Superpuppy 3
K6-3+ 450 ACZ (6x100)
DFI K6BV3+/66 Rev B2 (2 Meg) w/ 2x28mm Chipset Fans
2x256 Meg PC 133 Hynix SDRAM
1x 20G Maxtor (7200)
2x 80G Maxtor (7200) Ducted w/ 2x486 Fans Mount
52/24/52/16 LG CDR/RW/DVD
8/4/3/12/24/16/32 LG Super Multi
ATI 9000 aiw Radeon AGP
SB Audigy 1 MP3 Sound
CMD 649 IDE Controller
NEC USB 2 Card
K6-3+ 450 ACZ (6x100)
DFI K6BV3+/66 Rev B2 (2 Meg) w/ 2x28mm Chipset Fans
2x256 Meg PC 133 Hynix SDRAM
1x 20G Maxtor (7200)
2x 80G Maxtor (7200) Ducted w/ 2x486 Fans Mount
52/24/52/16 LG CDR/RW/DVD
8/4/3/12/24/16/32 LG Super Multi
ATI 9000 aiw Radeon AGP
SB Audigy 1 MP3 Sound
CMD 649 IDE Controller
NEC USB 2 Card
as the board has no jumper settings for setting fsb, voltage or multiplier, the bios boots at default settings. that means @ 2,2V and multi 2x, which is not remapped to 6x.
the sdram i used is from apacer - pc133 cl2, that means pc143 cl3. these sticks are also able to run 150mhz @ cl3, as i dont noticed any problems doing so.
now to wpcredit - its nearly impossible to find the specs for the sis540. what you can do is comparing the sis530 and sis630 chipset specs for finding similarities and assuming them for yours. i was not doing a lot of tests tweaking my system, because i was just trying to reach higher fsb speeds.
if the system is not stable, there is no goal using wpcredit.
thats why nearly no wpcredit test were done. but as far as i can say, the sis630 specs are a bit usefull.
the sdram i used is from apacer - pc133 cl2, that means pc143 cl3. these sticks are also able to run 150mhz @ cl3, as i dont noticed any problems doing so.
now to wpcredit - its nearly impossible to find the specs for the sis540. what you can do is comparing the sis530 and sis630 chipset specs for finding similarities and assuming them for yours. i was not doing a lot of tests tweaking my system, because i was just trying to reach higher fsb speeds.
if the system is not stable, there is no goal using wpcredit.
thats why nearly no wpcredit test were done. but as far as i can say, the sis630 specs are a bit usefull.
Are you sure it is at 2.2v ? That would be default for some K6-2s But not for all of them and not for K6-(2or3)+s. Does the bios have any alternate voltage settings ? If not could some be added by way of the bios patcher ? If you are using a K6-2+ and the bios is recognizing it as such, it might be setting the voltage at 2.0v. which would certainly limit your OCing stability.
EDIT : Sometimes you can use WPCredit to enhance stability by toneing down some of the settings. Takes experimentation for that.
EDIT : Sometimes you can use WPCredit to enhance stability by toneing down some of the settings. Takes experimentation for that.
Superpuppy 3
K6-3+ 450 ACZ (6x100)
DFI K6BV3+/66 Rev B2 (2 Meg) w/ 2x28mm Chipset Fans
2x256 Meg PC 133 Hynix SDRAM
1x 20G Maxtor (7200)
2x 80G Maxtor (7200) Ducted w/ 2x486 Fans Mount
52/24/52/16 LG CDR/RW/DVD
8/4/3/12/24/16/32 LG Super Multi
ATI 9000 aiw Radeon AGP
SB Audigy 1 MP3 Sound
CMD 649 IDE Controller
NEC USB 2 Card
K6-3+ 450 ACZ (6x100)
DFI K6BV3+/66 Rev B2 (2 Meg) w/ 2x28mm Chipset Fans
2x256 Meg PC 133 Hynix SDRAM
1x 20G Maxtor (7200)
2x 80G Maxtor (7200) Ducted w/ 2x486 Fans Mount
52/24/52/16 LG CDR/RW/DVD
8/4/3/12/24/16/32 LG Super Multi
ATI 9000 aiw Radeon AGP
SB Audigy 1 MP3 Sound
CMD 649 IDE Controller
NEC USB 2 Card
yup,
2,2V and amd k6-2+. exactly that tells me the bios and everest. as i used the bios for a winfast board, which realy has jumpers for voltage setting, no such option is implemented in the bios.
but voltage is not the problen, jim. the cpu boots up @ the 2x multiplier, so with the maximum 166mhz fsb i would get 333mhz cpu frequency. even 1,3v would be enough for that.
the disadvantage of wpcredit - u dont know, where to begin! without a pcr it is comparable to child in a jumbo cockpit. i simply dont have the time trying out every button to look for worse or better instability!
2,2V and amd k6-2+. exactly that tells me the bios and everest. as i used the bios for a winfast board, which realy has jumpers for voltage setting, no such option is implemented in the bios.
but voltage is not the problen, jim. the cpu boots up @ the 2x multiplier, so with the maximum 166mhz fsb i would get 333mhz cpu frequency. even 1,3v would be enough for that.
the disadvantage of wpcredit - u dont know, where to begin! without a pcr it is comparable to child in a jumbo cockpit. i simply dont have the time trying out every button to look for worse or better instability!
- KachiWachi
- K6'er Elite
- Posts: 507
- Joined: Wed Sep 21, 2005 10:53 am
- Location: Pennsylvania, USA
I would go back to the original BIOS for this board.
Moderator - Wim's BIOS
PC #1 - DFI 586IPVG, K6-2/+ 450 (Cyrix MII 433), 128 MB EDO. BIOS patched by Jan Steunebrink.
PC #2 - Amptron PM-7900 (M520), i200 non-MMX, 128 MB EDO
PC #3 - HP8766C, PIII-667, 768 MB SDRAM
PC #4 - ASUS P3V4X, PIII-733, 256 MB SDRAM
PC #5 - Gateway 700X, P4-2.0 GHz, 768 MB PC800 RDRAM
PC #6 - COMPAQ Evo N1020v laptop, P4-2.4 GHz, 1 GB PC2700 DDR
PC #7 - Dell Dimension 4600i, P4-2.8 GHz, 512 MB PC2700 DDR
PC #8 - Acer EeePC netbook, Atom N270 @ 1.60 GHz, 1 GB RAM
PC #9 - ???
PC #1 - DFI 586IPVG, K6-2/+ 450 (Cyrix MII 433), 128 MB EDO. BIOS patched by Jan Steunebrink.
PC #2 - Amptron PM-7900 (M520), i200 non-MMX, 128 MB EDO
PC #3 - HP8766C, PIII-667, 768 MB SDRAM
PC #4 - ASUS P3V4X, PIII-733, 256 MB SDRAM
PC #5 - Gateway 700X, P4-2.0 GHz, 768 MB PC800 RDRAM
PC #6 - COMPAQ Evo N1020v laptop, P4-2.4 GHz, 1 GB PC2700 DDR
PC #7 - Dell Dimension 4600i, P4-2.8 GHz, 512 MB PC2700 DDR
PC #8 - Acer EeePC netbook, Atom N270 @ 1.60 GHz, 1 GB RAM
PC #9 - ???

this actually has nothing to do with sis540, but :
180mhz sdram clock. i was impressed:
http://www.oc.com.tw/article/0103/readg ... asp?id=301
use the google implementation for translating chinese. seem to be very rare those tonicom modules.
180mhz sdram clock. i was impressed:
http://www.oc.com.tw/article/0103/readg ... asp?id=301
use the google implementation for translating chinese. seem to be very rare those tonicom modules.