CPU/Mobo Foxconn NF4XK8MC Mobo 1T Issue

Status
Not open for further replies.

dipdude

Forerunner
deejay, i have the same mobo, yes 1T is a strict no no, i am doing 2250 mhz, ht*4, htt*250, mem(5:6) 2.5,3,3,7 @ 2T, prime & memtest stable ;)

ok, just all the posts, yes my temp's too start (bios & ntune) @ 45`c & go upto 60`c on load, looks like mobo glitch but not sure what do you think ? do post a screenie using ntune.



with everest the temp's go haywire from 14`c - 60`c

 
@ dipdude,
To enable 1T use A64tweaker program and do check the benchmarks whether there is indeed any memory bandwidth gain.

I reckon foxconn played it safe by disabling 1T in the bios so that the m/b is more compatible with "ordinary" memories. and i believe they may enable that option in future bios updates. (although the entries for selecting 1T are present in bios , enabling it renders the m/c unbootable.)

RE: temps , i had tried using the the latest nforce chipset drivers V6.67, especially the smbus drivers but now in everest no temps are reported except for the HDD !!!
ah well, back to playing around with the smbus drivers .....
have you tried the superstep program included with the superutilities package ..?? there the temps shown are fairly stable.

ps: what did the m/b cost you.??
 
i am able to boot with 1T only problem is it is not memtest stable, that's why i am sticking to 2T, but with your ocz it should be able to do 1T(maybe at 333mhz).

there is slight gain in memory bandwidth, with cnq it is 6667 mb/s & without it is 6545 mb/s.

as far as temp's concerned - superstep, bios, ntune report the same stable temp's, but always start from 45'c (that is why i had a doubt).

as for price i got the package(foxconn+6200tc) for about 6k before 4-5 months.

the latest nforce drivers should be 7.13 or 7.15 whql


also did you set your mem settings on manual ?
 
@ dipdude,
well found out the solution for the 1T problem. it was the max asyn latency value under it. no matter what relaxed settings i had set it to 7, or 8, or 9 it would not boot.
so i did the next logical thing and set the command rate to 1T and set the user config value back to "AUTO" instead of "manual". i left the max async latency and read pre-amble values as it is.
now the m/c boots with 1T settings and it is shown as such in the post screen.

in "Read" test in everest now i get score of around 6400 (Previous it was around 6150)

here is the screenie you wanted.



and that above link you posted, is it for amd64 platform ?? or is it for intel ??
 
so i did the next logical thing and set the command rate to 1T and set the user config value back to "AUTO" instead of "manual". i left the max async latency and read pre-amble values as it is.

yes i had to do the same thing, strange mobo how you set the value manually & change to auto, that's why i asked if you changed the setting to manual ;) but you found the solution on your own :)

my bad about the link, here's the link for 7.13

In read test i get 5363 mb/s :(

p.s - is it prime, memtest stable ...
 
Running the user config mode on manual prevented the m/c from booting i had known since morning.(had to reset the cmos 3 times because of it)

but the strange thing is that now on a normal boot with 1T and then subsequently running a64tweaker to check the values i get read pre amble and max async latency values of 5.5 and 7 respectively.


But if i set these or even relaxed settings in bios and then leave it on "manual" the m/c will not boot.
will try changing the values in bios and then revert back to "auto" mode for user config settings .......;)

dipdude said:
p.s - is it prime, memtest stable ...

i have not tested it yet......will do so later (installing h/s etc...) but the m/c is running without any problems since morning ...and about 30 ~ 40 software reboots.
 
but the strange thing is that now on a normal boot with 1T and then subsequently running a64tweaker to check the values i get read pre amble and max async latency values of 5.5 and 7 respectively.
deejay this is what happened with me, i used manual setting for 2.5,3,3,7 as hynix's spd is 3,3,3,8. but had loads of errors or would not boot even on >220mhz. then using same setting 2.5,3,3,7 i just put user config to auto, it worked.

the reason i noticed using rmma was that my tRC & tRFC were on 24 something on manual, but by setting the values & selecting auto made the tRC & tRFC to 10 & 12 resp. did not check my max async or preamble as i did not have a64 then.

When we select manual, the setting's selectable (tcl,trcd,trp,tras) change but others not selectable are set to some ridiculous values, thats why it messes up :(

so just by changing the values & putting to auto helped me.

@deejay, yes my hynix hates 1T, loads of memtest errors. glad you would push your foxconn further but i thought you would go for dfi soon.
 
Last edited by a moderator:
@ dipdude, No i am not going to be using this foxconn m/b. might sell / give it away soon (not because it has any drawbacks etc. infact it is a pretty stable board, being running here since yesterday morning non-stop without a single glitch)....it will be DFI for sure. its just that i got this m/b at a reasonable price.(about half the price you paid for yours approx.) and till i get the dfi i am in the "Value for Money" phase :no: hence testing out various cheap m/b's and actually posting the "field reports" here which may or may not be beneficial to many young members here who are currently not fortunate enough to spend a fortune on popular o/c toys.

ps: i know of a member who recently bought a m/b which costs as much as my this total rig (excluding the opty cpu) including the xfx 6600 gfx card.!!!!;) :no: :)
 
deejay a new bios has come out for the foxconn which fixes the temp bug, my idle temp hovers around 28-29c and under load maxes out 41-42c

here's the link to the bios, it is dated dec 07, 2005, the version is 133 (4BTW1P33.BIN) only thing is change the filename in flash.bat file to 4BTW1P33.BIN :) foxconn has messed up the version number ;)

here's a screen of temp's under stress of toast :
 
Status
Not open for further replies.