April 18, 20178 yr Crew Instead of pushing higher frequencies and looser timings, opt for tighter timings if you want real performance!!
April 28, 20178 yr 3200 at 100 strap or 3333MHz (125 strap) 12-12-12-21 at 1.55 should be doable if your IMC/mobo can handle it... Which board mate ? MSI Godlike/X99 SOC Champion here only do 3200MHz 32M stable, R5E/Godlike Carbon does 3333MHz with same hardware. SA voltage is important!! [ATTACH]5557[/ATTACH] [ATTACH]5558[/ATTACH] [ATTACH]5559[/ATTACH] Hey mate I have been testing my new cpu and R5E but I have found it to be very sensitive. I have been using xtu as a base comparison score, and to compete in a local ranking, but as soon as I tighten my ram at all xtu fails. If I go tighter than 16-16-16 I get error with no score. If I run 1T at any speed I get the same. I thought xtu bench was broken as I couldnt get a score at any settings, then set back to xmp (3200-16-17-17-38-2T) and I get a score again. Is this just an issue with xtu? Or is it the first thing to show instability?Odd thing is xtu stress test passes at the same settings that the xtu bench fails ... Any help would be great.
April 30, 20178 yr Yes I was using the latest version but with all the failed attempts I googled and they said 5.xx was stable, I installed that and its exactly the same. I reset to xmp or close to it and it passes. Anything tighter than cas16 or 1T and it fails (no score). What else would you guys advise is a good stability tester? I think a few hours of BF1 is the best test but I dont want to do that.
April 30, 20178 yr Crew superpi 32M, if that passes all should pass, normally not needed on X99 maybe try to set maxmem. Are you running a 32 or 64 bit OS ?
April 30, 20178 yr W10 64bit I was using xtu as a quick efficiency (and stability) check, what else can I use? I find that changes often affect 1m and 32m differently and I get sizable variations in 1m runs so it seems very inconsistent. Do you guys still measure mem bandwidth? I think I will test my timings in CB as it seems to respond to ram changes well. Thanks.
May 1, 20178 yr Crew try at least 16 or 32MB for memory stability, most stable OS for reliable is XP there, win10 results can vary... Anyhow if it passes one of the two former it is stable... I'm trying to finish a small article on it.. I included some more info in the article at : MadShrimps Consider it as starting point. a lot will depend on your IMC quality and the board's bios options Edited May 4, 20178 yr by Leeghoofd
May 10, 20178 yr I have got it to pass 32m down to 13-13-13-35-2T but scores seem to drop a bit and BF1 crashes almost instantly, I dropped back to 14-14-14-36 and its better but BF1 still crashes eventually, so safe 15-15-15 can play for as long as you like and no crashes. These are all 2T, as soon as I try 1T I load windows fine but benchmarks crash or error out. I havent tried higher than 1.55v yet.
May 13, 20178 yr New 4333 2x8gb kit for $624, I assume this is B-die. Has anyone gotten hands on? http://www.corsair.com/en-us/vengeance-lpx-16gb-2x8gb-ddr4-dram-4333mhz-c19-memory-kit-red-cmk16gx4m2e4333c19r
June 10, 20177 yr Doing what this b-die can do for daily, surprised me DDR4 4333 18-19-19-39 2T 1.4v (Manual Sub Timing for Performance Boost, Auto sub timing have very low performance).
June 11, 20177 yr A lot cheaper at Newegg: https://www.newegg.com/Product/ProductList.aspx?Submit=ENE&IsNodeId=1&N=100007611%20601298182
June 23, 20177 yr 3200 at 100 strap or 3333MHz (125 strap) 12-12-12-21 at 1.55 should be doable if your IMC/mobo can handle it... Which board mate ? MSI Godlike/X99 SOC Champion here only do 3200MHz 32M stable, R5E/Godlike Carbon does 3333MHz with same hardware. SA voltage is important!! [ATTACH]5557[/ATTACH] [ATTACH]5558[/ATTACH] [ATTACH]5559[/ATTACH] I have the x99 soc champ capable of running 3466Mhz 12-12-12-28 With a 4x8Ghz set of 3200Mhz cl14 g.skill. This kit is not the best set either. Tested on z170M oc formula with a 7350k and was able to only get it superpi 32m stable @ 3466Mhz 12-12-12-28 using the 3600Mhz cl12 preset the oc formula has. I also have the 2x8GB 3600Mhz cl16 kit that will do 4000Mhz 12-11-11-28 @ 1.94v but have not tested in x99, looking for another good set. So im pretty sure the soc champ can do better than 3333Mhz with a good set of b-die ulness you were specifically talking about the zadak b-die http://forum.hwbot.org/attachment.php?attachmentid=5758&stc=1&d=1498206792 http://forum.hwbot.org/attachment.php?attachmentid=5759&stc=1&d=1498206792 Edited June 23, 20177 yr by lilchronic
June 23, 20177 yr is that 3466 pi32M stable ? Anyway you got a good combo rocking man!!! honestly i haven't tested 32m on x99 with those stick's that much as i don't want to push my chip that hard on just water cooling. My last 5820k i killed the imc some how and it would only run dual channel. When i tested those stick's on z170 thou i was only able to get 32M stable @ 3466Mhz 12-12-12-28 max with the two best stick's no matter what voltage i used i couldent get higher. Weird thing is i can take these stick's all the way up 4300Mhz with 19-21-21- and run a stable 3733Mhz 17-17-17-37-1T @ 1.4v and 500% of hci memtest. These are the 4x8Gb 3200Mhz 14-14-14-34 G.skill, My 3600Mhz cl16 kit is much better i just need another set that will match what they can do.
July 11, 20177 yr Hello The G.Skill 4000 C18 kits have B-Die of the same quality as the 3600 C15 kits? Or it is rather the same quality as the 3600 C16 ? Just to know if with a 4000 C18 kit I can have 3600 C15 to 1.35v Thanks
July 14, 20177 yr Hello The G.Skill 4000 C18 kits have B-Die of the same quality as the 3600 C15 kits? Or it is rather the same quality as the 3600 C16 ? Just to know if with a 4000 C18 kit I can have 3600 C15 to 1.35v Thanks Up please To know if I take the kit 4000 C18 and I can have 3600 C15 to ~ 1.35v (and after a higher frequency on X299) Or if I take the kit 3600 C15 (stable at 3733 C16 and 1.35v) Thanks so much
July 14, 20177 yr Up please To know if I take the kit 4000 C18 and I can have 3600 C15 to ~ 1.35v (and after a higher frequency on X299) Or if I take the kit 3600 C15 (stable at 3733 C16 and 1.35v) Thanks so much Ok thanks, so with a kit G.Skill 4000 C18, it is possible to have 3600 C15 to 1.35v G.Skill only made 1 kit 3600 C15 (no 3200 C13 for example) at 1.35v But many kit 3200 C14 or 3600 C16 And I was wondering if the 4000 C18 kits had the same B-Die as the 3600 C15 or rather the same as the 3200 C14, 3600 C16 But to be sure, will have to test I think I saw a review where with the 4000 C19, it was 3200 C15 at 1.35v I guess with 4000 C18, it's 3400 C14 and so impossible to have 3600 C15 Edited July 14, 20177 yr by tistou77
July 15, 20177 yr With a kit G.Skill 3200 C14, the 4000 C18 is "stable" at 1.35v No need I test kit 4000 C18 I think, I could not have 3600 C15 (or 3733 C16, eg) at 1.35v
August 3, 20177 yr Could someone confirm if the Corsair Dominator 3466 C16 CMD16GX4M2B3466C16 is using the Samsung B die?
August 3, 20177 yr Could someone confirm if the Corsair Dominator 3466 C16 CMD16GX4M2B3466C16 is using the Samsung B die?Unfortunately 3466 kits with 8GB DIMMs can either be single-sided B-die or double-sided E-die. Sent from my SM-G955U using Tapatalk
August 15, 20177 yr Try G.SKILL Trident Z 3466 C16-18-18 from Quad Channel kit, this is two best stick. But still no try waza, waza looks like stress my IMC much with my CPU on ambient. This is at 2.035v not try less :
August 16, 20177 yr Continue test 3466 c16 kit, can run 4133 12-12-12 2.05v (no waza), 4090 12-12-12 2.035v (waza). All air, still need to find the limit...
Join the conversation
You can post now and register later. If you have an account, sign in now to post with your account.