February 19, 20169 yr Administrators Can be IMC or memory, iyou can try to vary mamem or waza size, if it doesn´t help it will be hardware related
March 1, 20169 yr Any tips for booting Samsung B-die at 4000 on M8E? I tested a 6600k and 6700k with some LPX 3600C18, on Asrock with either CPU all sticks can boot into XP but on Asus, the board halts Code 3F no matter which CPU. Sent from my SM-G928T using Tapatalk
March 1, 20169 yr Author Are you trying to run with 1T ? If so this is the problem,normal boards will give you a wall at 3900-3950 with 1T,you need an Impact for 4000+1T You can try to use my timings from here: This was very easy to use 3866 divider then raise BCLK , if you use 2T you should have no problem for 4000,i will test that too later today.
March 1, 20169 yr Definitely 2T, I did attempt 1T but that wasn't working on either board. 2T boots fine on Asrock with 100BCLK, is the 4000MHz divider bugged on Asus?
March 1, 20169 yr Administrators No, the 4000 and 4133 dividers work - I ran 4000 1T with E-die, and higher, 32m on Hero. B-Die didn´t work. It seems all 4 Dimm boards have a hard time to handle 4000 divider with 8GB sticks even when the sticks are single sided Edited March 1, 20169 yr by websmile
March 1, 20169 yr No, the 4000 and 4133 dividers work - I ran 4000 1T with E-die, and higher, 32m on Hero. B-Die didn´t work. It seems all 4 Dimm boards have a hard time to handle 4000 divider with 8GB sticks even when the sticks are single sided Yeah I forgot E-Die works, but as far as 4 DIMM boards and B-Die go, OCF has no issues 4GHz 2T. Sent from my SM-G928T using Tapatalk
March 1, 20169 yr Yeah I forgot E-Die works, but as far as 4 DIMM boards and B-Die go, OCF has no issues 4GHz 2T. Sent from my SM-G928T using Tapatalk Asus has the T-topology. Which helps them with 4 stick efficiency, but cripples their 2 DIMM max freq, a little bit.
March 1, 20169 yr Author For avoiding 3f set twrrd_sg and dg manually, you can use 22 19 for startup at c12, this will go lower or higher with cas Sent from my SM-N910F using Tapatalk
March 1, 20169 yr For avoiding 3f set twrrd_sg and dg manually, you can use 22 19 for startup at c12, this will go lower or higher with cas Sent from my SM-N910F using Tapatalk I didn't even consider those values. I'll start with what Asrock sets and go from there. Sent from my SM-G928T using Tapatalk
March 2, 20169 yr Well now I just get code 41, i guess that's progress.[emoji19] [EDIT] I still had 1T set, 2T boots! Now to see if I can load Windows without crashing. Sent from my SM-G928T using Tapatalk Edited March 2, 20169 yr by sabishiihito
March 2, 20169 yr Author Glad it helped! Sent from my SM-N910F using Tapatalk Edited March 2, 20169 yr by Alex@ro
March 5, 20169 yr Oddly enough I can't seem to replicate this on the Extreme board with the exact same settings. Sent from my SM-G928T using Tapatalk
March 6, 20169 yr 4000Mhz seems impressive , but the PI 32M time is so slow... Well no wazza and Asus 4-DIMM boards aren't supposed to be that efficient with only 2 sticks, right?
March 7, 20169 yr Crew Well I ran Win 8.1 64bit OS, 2 sticks at 3733C12, CPU at 4Ghz and had a time of 5 sec less...no wazaa, no nothing XP should be way faster for Pi
March 7, 20169 yr Author Runs without wazza can vary much Albrecht Sent from my SM-N910F using Tapatalk
March 7, 20169 yr Crew I know but I expected way more than a mere 2 secs on a similar platform, biggest difference should be XP versus 8.1, a bunnyextractionload more volts, higher mem speeds,...
March 15, 20169 yr Impact is absolutely outstanding with B-die. All air, and only one really good stick, still haven't been able to find a good match for it.
March 15, 20169 yr Pm me what vdimm that stick was for 3733c12 on any big board. (Actual). I'll see if I've got one to match it for you.
April 3, 20169 yr Finally got 2KC12 1T to pass on my new M8I. Mismatched sticks, but I only had 3 out of 12 B-die modules that would pass 32M at these settings and one needed more voltage than these two could be stable at.
April 4, 20169 yr Finally got 2KC12 1T to pass on my new M8I. Mismatched sticks, but I only had 3 out of 12 B-die modules that would pass 32M at these settings and one needed more voltage than these two could be stable at. You should be locking in them rtl's after initial training, your first stick is at 58 rtl and 15 iol where on Impact 4000/12 should be 49/50/6/6 or possibly even lower iol
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.