Huge Problems with 1:1 Divider on the Foxconn Mars

[mpc]

Hi!

I was just trying some things with my new Foxconn Mars and my E8200.

I was trying to get the FSB over 535 Mhz, but nothing worked out. I guess finally managed to localize the Problem.

Every Time I put the Divider to 1:1 I can't get the system to run through SPI over 535 Mhz. It dies all the time with Bluescreens.

When I use the 2:3 or 4:5 divider, the ram is running through prime stability test with 600Mhz at 5-5-5-15 with 2.1V.

The 8200 is pretested up to 4800Mhz (FSB 600), so there is no problem either.


I'm using the P12 Bios. The Voltages are set to:

CPU: 1.5V (Voltage needed for 600Mhz FSB/4800Mhz)
Ram: 2.3V Timings: 6-6-6-18 2T Divider: 1:1
NB: 1.5V
VTT: 1,35V
SB: 1.613V

Now I will try to use some differents Bios.

Has anybody an idea what to do? Somebody with similar problems?

I know I already asked for help at the Quantom Force Forum, but its down all the times, so I'm not really able to watch the thread there!

[mpc]

Update:
Updating the Bios to D16 and G30 isn't changing anything

[LardArse]

For high fsb + high mem speed push more mch voltage, 1.7v+
but at the end 550+ fsb usually you wont be able to run that high a memory speed.

[mpc]

Thanks for the answer Shamino

The Problem is that 1:1 is the smallest divider I can get. There is no option like 2:1.

So is there anyway I can get close to 600Mhz FSB?

[LardArse]

4:5 divider, cl5-5-5-15 2t, try adjusting performance level looser with memset, 1.7v nb see if u can go higher

[mpc]

Ok I will try that. Thanks for help

[saaya]

I know I already asked for help at the Quantom Force Forum, but its down all the times, so I'm not really able to watch the thread there!

thats why we are over here at xs now

did you play with the gtl settings?
what cpu multi are you running?
how are you cooling the nb?
what memory slots are you using?

[mpc]

@Shamino:

I can go up to 615Mhz on the ram, with 618Mhz Spi will only run till loop 15 with these settings...

@ sayaa:

What Gtl settings? Didn't know that the Mars has some?!
I tried 6 and 9, neither one works!
I'm cooling the NB with the stock heatpipe and the Stock fan @ 100% speed!
I tried slo 1/3 and 2/4, nothing will change -.-

Thanks for help guys!

[SoF]

Maybe I can join here just not to open another topic about mars (P12 bios currently):
I have E8400 running here now on Mars and it tops out at ~520 FSB (x6, x7, x8). I started to test RAM (it is an older DDR2 Cellshock kit from Sayaa which did 600 4-4-4 on other boards.
I ended up with 550 max 4-4-4 (2.45V) and 580 5-5-5 (after 32m system crashed ^^).
I went on testing a lot with sub timings but nothing helped. I am waiting for my E6850 to get back so I can check if the chip is crap. High nb voltage seems to help a lot past 510 FSB but I am a bit wondering that it takes that much of it. I will test it under dice once I guess (when I get it tomorrow) and then change to P45
Any tipps for the sub timings on D9GMH with 580+? Tried a lot but seems that a lot does only help with superpi speed but not stability.

[saaya]

hey sof! good to see you around
too bad i couldnt make it to that BBQ say hi to your gf from me hehe
i havent played with the mars at all tbh
ill ask shamino if he can give you guys some more tips

[SoF]

yeah what a pitty but maybe during GC (it is only 160km from here to leipzig and when you fly via Berlin...) we might have the possibility to catch up our BBQ
maybe shamino knows what does the trick, the FSB I say is related to the CPU but the RAMs should do a bit better. only tested P12 bios until now as I can't read of improvements with further beta bios or similar.
ah additionally one further question: is it possible that the bios sometimes doesn't recognize the subtimings in bios. e.g. I set TRRD 4 and in windows memset says trrd 8 - my feeling is that this happens quite often.

[aoch88]

I guess Foxconn is not giving much support for the MARS anymore. Btw, what's the change log for BIOS D16?

[saaya]

yeah what a pitty but maybe during GC (it is only 160km from here to leipzig and when you fly via Berlin...) we might have the possibility to catch up our BBQ
maybe shamino knows what does the trick, the FSB I say is related to the CPU but the RAMs should do a bit better. only tested P12 bios until now as I can't read of improvements with further beta bios or similar.
ah additionally one further question: is it possible that the bios sometimes doesn't recognize the subtimings in bios. e.g. I set TRRD 4 and in windows memset says trrd 8 - my feeling is that this happens quite often.

hmmm afaik the timings in memset and bios tend to differ cause the timings are all slightly dependant on each other... and they are defined different by different chipsets. i talked to felix, the programmer of memset and he explained that some timings are a combination of X and Y, and while you can select Y in BIOS, the cant read Y in windows, he can only read the result of X plus Y. when you set 4 in bios does memset say 9? or does it say 10?

ive noticed both, offset differences and double offset differences from what you set in bios and what memset reports.

I guess Foxconn is not giving much support for the MARS anymore

why do you think so?

Btw, what's the change log for BIOS D16?

let me check, but i dont think ill be able to get it since its been a while...
why do you ask?

[mpc]

@SOF: Sounds pretty much like my Problem -.-
Did you try the E8400Max with the 1:1 Multiplikator? Around 520, that's when my system starts to get unstable, sounds like we have the same probleme here -.-

Where exactly are you from, its only about 120km from me to Leipzig

[saaya]

hmmm sounds like some BIOS limitationat around 520fsb then? hmmmmm

[mpc]

Yeah But using a different Bios isn't changing anything. I tried them All!

[SoF]

@SOF: Sounds pretty much like my Problem -.-
Did you try the E8400Max with the 1:1 Multiplikator? Around 520, that's when my system starts to get unstable, sounds like we have the same probleme here -.-

Where exactly are you from, its only about 120km from me to Leipzig

from south of berlin

1:1 highest FSB today with 1.65V NB and DICE on CPU


535 FSB seems to be benchstable for 32m and 3D until now - that was my last test until now with 1.63V NB. Still have 15kg DICE left so still some time to play

[mpc]

Yeah but that's under Dice...I want to use that FSB 24/7 -.-

I also answered at the Forumdeluxx Mars Thread

[SoF]

after some time benching 530 FSB board became unstable and I am now back to the FSB I had on air (515-520)...the strange thing is I can bench everything at 515 and it will give my a bsod or reboot with 2 mhz higher. hope to get my good old E6850 back tomorrow so that I can check if it is just the chip maybe.
535 FSB 32m in the afternoon but since some time this will only give me reboot or bsod now

[mpc]

The point is, that I have the same issues with an E8200 that is approved to run 600Mhz FSb and 4800Mhz without any problems. And the problems I have start around the same FSB range, where your problems start.

[SoF]

yeah that's very strange. At least your ram does 600 mine won't do that unfortunatly. I have to test my gskill CL5 kit maybe (627 5-5-5 tested) the cellshock does not like cl5 but cl4 seems to be even more unstable at higher clocks on the mars so maybe its just not the right kit for that board.

[aoch88]

let me check, but i dont think ill be able to get it since its been a while...
why do you ask?

If it's only some small changes, I won't bother to update this BIOS and I'm currently on P12.

[aoch88]

Btw in the "Memory TIming" menu, what does "memory channel dll adjust" do? Is it the same as those GTL reference voltage?

[mpc]

Ok, try that other Ram kit and tell me how that worked out!

[mpc]

Now 612Mhz on Ram are running through SPi without any problems!