Guest Bullant Posted October 7, 2015 Share Posted October 7, 2015 Since I'm in this thread already, bought the board yesterday - first result with Samsung, quite easy to get a start using the presets Will get XP on there ASAP. Nice work man Quote Link to comment Share on other sites More sharing options...
Guest Bullant Posted October 7, 2015 Share Posted October 7, 2015 XP won´t be faster without waza anyway, you won´t save time . @Luumi I made a couple of full-slot tests, this works but is not too easy to handle, binning should be done in single or dual when you think one stick is weak, all air, so results will be easier with wc or cold Very hard on skylake to keep good real tight rtls on air for me, no matter which board I use or mems, IMC on air fights hard These are mfr web? @Luumi looks like it's bad dim mate if it's failing at those low volts,all the other dims keep scaling 1.65v? Quote Link to comment Share on other sites More sharing options...
Guest Sinebrychof Posted October 7, 2015 Share Posted October 7, 2015 (edited) They scaled up to 1.65v at 3000 cl11 yes, but like splave mentioned in the old MFR thread at oc.guide the higher voltage starts to do more bad than good when going near 3200. I actually havent managed to post 3200. To get near 3200 I need to post and boot like 3150 and then set in OS. Are correct RTL/IOls at 3200 cl12 46-47/6-6 or 47-48/7-7 without tweaking the IOls? Edited October 7, 2015 by Sinebrychof Quote Link to comment Share on other sites More sharing options...
l0ud_sil3nc3 Posted October 7, 2015 Share Posted October 7, 2015 Since I'm in this thread already, bought the board yesterday - first result with Samsung, quite easy to get a start using the presets Will get XP on there ASAP. Glad to see my 4000 c15 tight profile is working well As Bully said man, awesome work 4000 c15 first touch @ air is a phenomenal start, 4200 c16 or bust next Quote Link to comment Share on other sites More sharing options...
Guest Bullant Posted October 7, 2015 Share Posted October 7, 2015 They scaled up to 1.65v at 3000 cl11 yes, but like splave mentioned in the old MFR thread at oc.guide the higher voltage starts to do more bad than good when going near 3200. I actually havent managed to post 3200. To get near 3200 I need to post and boot like 3150 and then set in OS. Are correct RTL/IOls at 3200 cl12 46-47/6-6 or 47-48/7-7 without tweaking the IOls? 47-48 rtls are fine for cl12 Quote Link to comment Share on other sites More sharing options...
Administrators websmile Posted October 7, 2015 Administrators Share Posted October 7, 2015 These are mfr web? @Luumi looks like it's bad dim mate if it's failing at those low volts,all the other dims keep scaling 1.65v? Yes, MFR Quote Link to comment Share on other sites More sharing options...
aerotracks Posted October 7, 2015 Share Posted October 7, 2015 (edited) Glad to see my 4000 c15 tight profile is working well As Bully said man, awesome work 4000 c15 first touch @ air is a phenomenal start, 4200 c16 or bust next Hey loud, working all smooth, thanks for sharing Just started single stick testing, this plus spending lots of time in the UEFI and seeing how training behaves goes a long way. Will attempt C16 (and fail, probably) - however I have a feeling I'm not done with C15 just yet Edit: 4100C15 Dual Edited October 8, 2015 by aerotracks Quote Link to comment Share on other sites More sharing options...
aerotracks Posted October 8, 2015 Share Posted October 8, 2015 Tightened them down to CAS11 Quote Link to comment Share on other sites More sharing options...
Alex@ro Posted October 8, 2015 Share Posted October 8, 2015 Impressive,how many volts and what model do you have? Quote Link to comment Share on other sites More sharing options...
Administrators websmile Posted October 8, 2015 Administrators Share Posted October 8, 2015 First test on new cpu and new mems and final test on board - maybe I will get used to sams at my tests, more fun for me than hynix which I did not fancy on ddr3, we all have our preferences^^ Let´s keep on quizzing Quote Link to comment Share on other sites More sharing options...
NickShih Posted October 9, 2015 Share Posted October 9, 2015 Tightened them down to CAS11 This is impressive.. never try this before. 1.9 VDIMM? Great work Quote Link to comment Share on other sites More sharing options...
aerotracks Posted October 9, 2015 Share Posted October 9, 2015 (edited) Thanks guys, VDIMM little bit above 2V. Model is Team Xtreem 3866, only high bin Samsung I was able to find in online shops^ Edit: Let´s keep on quizzing Not bad Edited October 9, 2015 by aerotracks Quote Link to comment Share on other sites More sharing options...
Administrators websmile Posted October 9, 2015 Administrators Share Posted October 9, 2015 (edited) I guess my 3200 gskill I bought used will not compete against this^^ P.S. a bit above 1,6v real for my run, not sure how scaling is at high regions, mine are d die Edited October 9, 2015 by websmile Quote Link to comment Share on other sites More sharing options...
aerotracks Posted October 9, 2015 Share Posted October 9, 2015 (edited) One stick fails 20-20, so it's 11-21-21. Win7 didn't care, but in XP it matters. 4000C11 2.17V, small waza For some reason I can't train above 2010MHz with 1T, doesn't matter if it's C11, C15 or C19. Wonder what that's about. Edited October 9, 2015 by aerotracks Quote Link to comment Share on other sites More sharing options...
ObscureParadox_3XS Posted October 10, 2015 Share Posted October 10, 2015 This is with cold memory right? I've not been keeping track of DDR4 clocking at all but that speed seems way too high for air Quote Link to comment Share on other sites More sharing options...
Guest Bullant Posted October 10, 2015 Share Posted October 10, 2015 This is with cold memory right? I've not been keeping track of DDR4 clocking at all but that speed seems way too high for air This will be air mate,this new Samsung-E is really good,be interesting to see how it goes on cold,even if only cold it tightens trcd-tras it still be good It's is very high volts so where yet to see if it will degrade the memory that I've been told it can Quote Link to comment Share on other sites More sharing options...
ObscureParadox_3XS Posted October 10, 2015 Share Posted October 10, 2015 Well then, sign me up for some Samsung-E, I really want to get hold of some and have a go on cold with it now. Like you said even if we only see a little bit of scaling with it, it's going to be brilliant on cold Quote Link to comment Share on other sites More sharing options...
Guest Bullant Posted October 10, 2015 Share Posted October 10, 2015 Well then, sign me up for some Samsung-E, I really want to get hold of some and have a go on cold with it now. Like you said even if we only see a little bit of scaling with it, it's going to be brilliant on cold Someone might want to confirm but yeah I'm told it is the new sammy-E ,yeah it does look good and want a kit too Nice result to aero Quote Link to comment Share on other sites More sharing options...
NickShih Posted October 10, 2015 Share Posted October 10, 2015 One stick fails 20-20, so it's 11-21-21. Win7 didn't care, but in XP it matters. 4000C11 2.17V, small waza For some reason I can't train above 2010MHz with 1T, doesn't matter if it's C11, C15 or C19. Wonder what that's about. It is speed limit by IC maybe. Try to loose trp/trcd might help you Also the last 2 third timing. Quote Link to comment Share on other sites More sharing options...
rtsurfer Posted October 10, 2015 Share Posted October 10, 2015 NVM. Quote Link to comment Share on other sites More sharing options...
nickolp1974 Posted October 10, 2015 Share Posted October 10, 2015 i cannot get tcWL to 9 with this and not sure i dare push volts some more, at 1.94v here Quote Link to comment Share on other sites More sharing options...
aerotracks Posted October 10, 2015 Share Posted October 10, 2015 i cannot get tcWL to 9 with this and not sure i dare push volts some more, at 1.94v here Did you try relaxing tRDWR timings by one or two values? That's the first thing I would think of, I'll give this a try when I'm home. Quote Link to comment Share on other sites More sharing options...
aerotracks Posted October 11, 2015 Share Posted October 11, 2015 (edited) @nickolp1974 Hey so I tried 4000 13-21-21 last night, when I dialed in your timings it gave me a 55. I was able to do a non waza pass with tRDWR timings auto at 1.72V. Which memories do you have? Tightened it up a little at 4000C11. Long time effects of overvolting are a big guessing game at this point of course, seeing scaling with additional voltage is good enough in my book - should something die it'll have been sacrificed in the name of science Edited October 11, 2015 by aerotracks Quote Link to comment Share on other sites More sharing options...
nickolp1974 Posted October 11, 2015 Share Posted October 11, 2015 4000C13!! Nice any chance you could do a pi32 run at X45/45 multi so I can gauge perf. My mems are the team extreme 2x4gb 3866mhz kit which I presume are the Samsung-e. The max voltage I have tried so far is 1.97, in my short testing I'm finding the 3733 strap is working well for me which then allows a little headroom pushing up the bclk in a few bench's without going past my already uncomfortable voltage limit. Can't wait to see what Bullant can do with the samsungs, no doubt be putting everyone to shame!! Quote Link to comment Share on other sites More sharing options...
aerotracks Posted October 11, 2015 Share Posted October 11, 2015 Sure you're not heavily overvolting that C13 setting? I have the same kit, and as I've written, 1.72V for a non-waza 32M Pass. Quote Link to comment Share on other sites More sharing options...
Recommended Posts
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.