tWR 8 and tFAW 16 not possible bull?
Printable View
tWR 8 and tFAW 16 not possible bull?
Hey guys
@ C-N thanks man having a great time chipping away at my time,I'll keep trying
@ Splave I did try tFAW 16 and tWR 8 but it seems when I push things too tight my times become worse,also when I go tighter on tRFC 85 times get worse.This is only when I'm at the big 2700+ mhz tho,if I'm at lower Mhz it allows me to then go bit tighter.I'll keep trying them tho
l0ud_sil3nc3 - 6:17:500 - 104.2x48 - 1354.7MHz - 7-11-7-27-90-1 - Asrock Z77 OC Formula - G.Skill Flares 2000 c7/PSC - 2.0v - cpu LN2 (-65c) / memory LN2 (-94c)
http://i.imgur.com/aYGo5AT.png
Awesome run lOud, nice clocks that's the stuff
Had a play with this today 4x2Gb, no chance getting this stable.
http://s12.postimage.org/6ff8vrw65/image.png
not to be a d1ck but 1st place run does not meet the SS requirements :P
If you are referring to this
then you are correct, guess I never really noticed that:hehe:
only had an hour, so CPU on air. Not bad for 1300mhz only on the mems :P
http://i.imgur.com/6dZpSqa.jpg
seems new kit has potential, 2000 c6 pi
Will hook up SS tomorrow and try again
Managed to get the new KPC pot down to -120's Dimm temp :D about 50C colder then I could get the old revision down to.
http://i.imgur.com/Qpt7nus.jpg
^ beast as usual from you, the new cooler really is much better than prior designs at transferring the heat efficiently.
Very nice Splave
Hey guys any of you had a initial value of 4.232,I had this last night but crashed straight away
Yeah thanks Splave I think your correct, tried it tonight at 600 and it puts it under to much stress,so I tried one bit higher @ 690 and came back @ much the same as my 670.Maxmem 670 was 6.656 and maxmem 690 was 6.657.Im still just using one file moving it from D -C-D then run,may need to try a new waza setup
Higher maxmem than 700 would suck right?
I don't get how 4 dimms would benefit the score over two or even one...:confused:
I think its interleaving is the benefit
Article pretty interesting
https://www.google.com/url?sa=t&rct=...42553238,d.eWU
I'm gonna have to give it a shot...I have a lot of psc sticks to try anyway but it just doesn't add up with what I had previously known or thought worked best.
It helps :D
Yes it helps but it will stress your imc more and may result in a lower Mhz due to your imc can't handle it.What I mean by this is if you can reach 2666Mhz with 2 dims stable,you may not reach the same speeds with four dims stable.Four dim speed may reduce to say 2600Mhz stable due to a weak imc,even at the little lower speed tho 4 dims may still be faster
Yes, I'm aware of the imc limitations...I only have an fx8120 to test this stuff on atm. The imc is quite good but the mobo not so much (m5a88-v evo)
I'll be testing my psc stack with my new 3770k soon and post some results here.
Failed hard 4 dimms today, on ln2 couldnt even get 2600 7-11-7 and tight sec/thirds to go through 4 loops
Yep
http://i.imgur.com/eQZoQAw.jpg
finally hit a home run :D
http://i.imgur.com/H3NemuU.jpg
Splave - 6:16:547 - 49x102 - 1326.8MHz - 7-10-7-24-70-1 - ASROCK OC Formula - G.skill 2300c8 2x2gb/PSC - 2.025v - CPU SS @ -50c Mem LN2 @ -100c
2 DIMM FTW!
If you want to try my wazza, 35-36 copies of 01 in a folder. Single copy D->C
2 dimms... :p
01 as in...?
3dmark01.exe install file
Not bad alle, 3d01 ftw
nice result allen!
Keep pushing bro!
If i had a ivy setup i would give this a shot as I love 01 but i dont,unless someones wants to give me some cheap gear? :P
Very nice work Splave,awesome result
Figured it was 3dm01...:D
Cool, I should try doing some Bas Rutten vid wazza. That should help with the buttkicking part.
Dumo - 6:18:234 - 50x100 - 1400MHz - 9-12-12-21-107-1T - MVE - 4X4 Corsair Dommies 2666C10 - 2.1v - air/cpu ss
http://img849.imageshack.us/img849/7...2800twcl61.png
Uploaded with ImageShack.us
On MVE
Dram CP 2, TX slew ab 4, RX slew 5, MCH ds a 14, MCH ds b 15
No mercy for the weak :D
Had a little go again tonight,failed at loop 24,will try your waza out Splave.Seems im losing a lot on the last value output,with correct waza setup this should be in the 16.450s?
http://s4.postimage.org/pmdj2p7v1/fgg.png
So i was giving a shot at this...Day of failure for me and Bullant it seems :p:
Long story short... booted,set voltages,wazza'd and started to run pi...
OOOps no ram timings set so stop spi and memtweak the hell out of it and run straight again...
Oh run seems pretty good (my goal was below 6m17s) so decided to take a video of it why not...
Loop 24 was finished and shuddenly... kaboom system restart...
Decided to let it dry a bit and give it a shot an other time after i insulate the mem area a bit...
If wazza is done correctly maybe 6.16.500 is doable...
vid and mem settings:
http://www.youtube.com/watch?v=Fq5LzgpL1ts
http://img12.imageshack.us/img12/6691/screen002ga.jpg
Uploaded with ImageShack.us
Are you using ram disk? That will help your last loops.
Something is not right though I think with you're timings 4 dimm should be around 6:16.2**
Hey,im not using ram disc and my waza is just 1x1.8gb file that i move 2x.Will try ram disc,thanks for the help
@ crusty,awesome timings man,tell me crusty you wanted to swear on that fail lol
Totally man... i wanted something to begin with and start crunching it down... no worries though will try again (maybe try twcl6 to see what's happening)
are those PSC or BBSE, crusty?
Hey guys i just notice i haven't been running in debug mode under boot ini advanced options tab, If I tick debug mode it should help?I have themes enabled this is correct?
Ok now it's getting irritating... found some good wazza (big big file d-c 3 times) 5-6mb dif between available and cached...
it starts slow but then it cuts down really really fast and helps with bonus loop too...
dunno with i lost this one...was aiming for 6.16.4xx
http://www.youtube.com/watch?v=aFt6T...ature=youtu.be
bbse ln2 full pot 1.89v (no need for more) cpu is still aircooled maybe i will get it subzero next time...
Still working on my waza and no ram disk but debug mode may have helped,getting there will keep working on it.
bullant - 6:16:578 - 104.2x48 - 1354.3MHz - 7-11-7-26-90-1 - Gigabyte Z77X-UP5TH -G.Skill Pi 2200C7 4X2Gb/PSC - 2.15v -CPU SS -30 / Memory LN2 -110
http://s11.postimage.org/ghc3yt3lf/image.png
My secret manual wazza (after killing explorer, Pi opened and ready to go) :p
https://dl.dropbox.com/u/30576042/manual-wazza.JPG
l0ud, it doesnt really matter. You will notice fredyama's runs his available mem is at 477990912 and his score is great. My run is only 529838080, its more about the balance between them.
Yeah,but too high commit charge... if you wazza and wait a looooot in the end cached and available will become even but that is not the only key...
cas 6 and tWCL 8 is sllllllooowww
tFAW would not do 16 with CL6 for some reason either
http://i.imgur.com/lmAJZYL.jpg
Thanks for showing and testing Splave,did you test cas7 and tWCL8,was it even slower?
hey guys,
i can run 4x2Gb 1300 8-12-8 1.87V and my time is 6.20.xxx i didint came under this magic time.With this settings i must run easy 6.19.xx or faster but my copy waaaza ist not that good.Can you gave me some help to finding the rigt waaza
Hiwa - 6:18:568 - 50x100 - 1300MHz - 8-12-8-28-79-1T - MVE - 4X2 G.SKILL 16000PIS + GSKILL 16000Trident PSC- 1.95v - mem air/cpu ss -20
http://img189.imageshack.us/img189/3720/screen003le.jpg
LB: 5 ---24 X 2K1 D-C C-D 670MM
Are you guys using a ram disk besides LSC, maxmem and wazza?
Got 5 sticks of psc anyway...I gotta give it a try. And experiment with different wazza file sizes.
just checked this thread for the first time in a long time, amazing results angoholic, splave and bullant, keep pushing it guys, im sure we will see 16.3xx soon
Guys anyone have idea time at 4ghz ?
some BBSE run
http://img801.imageshack.us/img801/44/screen006q.jpg
@ hiwa,I can post my 4Ghz time up later and show you if you like its 43.609
@Pro,thanks mate
Here is two 4Ghz runs under cold,when i get a chance later will test again with tWCL6.Interesting the times tho from two different memory frequency speeds and boards
Gigabyte UD3H
http://s1.postimage.org/fl0lqac9b/5656.png
Gigabyte UP5TH
http://s10.postimage.org/tl0l3t4ux/cccfff.png
2 runs for today is enough
Ambient -10 CPU SS
http://img15.imageshack.us/img15/2769/screen007bi.jpg
just decide to set 2666 and passed with 2.05V
http://img803.imageshack.us/img803/407/screen009u.jpg
set to 2700 and passed 8 loops with 2.12V
whoooops hiwa nice bbse chips here :)
can you run the 1333 tight @ room temperature, too?
anyway nice, come get them some ln2 :)
Very nice BBSE, seems like yours will do very well with cold, if they can run that at -14
No need ln2 for 16s my ss -20 will be enough, just a cpu can run 5ghz air lol, this shift need -10 to pass 5ghz 1.47
:rofl: btw which kit do you use? from billy?
Some really impressive runs on air Hiwa
some first test Trident 1600CL6 BBSE @ 1.94V
cold air,ram HS temp 3.3C Grad
http://www.abload.de/thumb/trident16...008-1drs7n.jpg
i can run this settings on 5Ghz but with CW came not exact in round,too much for the imc or whatever
T0lsty - 6:17:203 - 50 x 100 - 1333mhz - 7-10-7-28 1T - Asus M5E - 2x2 G.Skill F3-1920CL9-4GBPIS PSC - 2.05V ; CPU -water /RAM LN2 -120.
https://lh3.googleusercontent.com/-K.../screen005.jpg
Messing with waza file sizes. . . 1.89GB D > C C>D
first no waza
http://i.imgur.com/ELGSaAl.png
waza
http://i.imgur.com/nUg6qrx.png
I don't know why the Asus MVF needs so much voltage to run compared to the OC Formula.
The above took 1.765v and 2600 8-12-8-28 with the same subs was not exact in round immediately, even though the same dimms can run 2600 c8 on the Asrock with TWCL at 7.
vccio 1.1
vccsa 0.96875
LB 5
just confusing to say the least.
This is nothing new that the OC Formula owns all Asus Boards on Vdimm, several kits I sold needed up to 0,1V less on OCF than on my Gene... In spite of MC being in cpu, the board is still a major factor, which some people tend to deny
Not only vdimm, OCF also takes 0.06v less vCore than MVE & MVF
MM or bios?
Vcore is true,however regarding vdimm,asrock ocformula overvolts by 0.025-0.027 on DMM,which is A LOT.I had an Oc formula,samsungs 2666C10 could not boot 2700 9-12-12 whenever on gene 2800 9-12-12 was 32M stable with 1.9V,don't know what to say about psc or bbse,overall i consider asus rog z77 boards to be the most memory-friendly....
On PSC, I have two reliable feedback including MM measuring, my gene undervolts with tool reading against bios by ~0,02, with mm ~0,03, Formula seems to overvolt, Alex is third one who confirms this, against Bios and tool reading ~0,03v, real differences I had were around 0,05v which GeneV needed more real voltage on the two psc kits than on OC Formula. As ususal, this can vary from board to board, but this is info I have so far.
hmmm, Mine reads exactly what I set. Strange
I think loud_silence has mm and can measure vdimm, after that hopefully at least his question will be solved
^ I don't mean it needs more voltage, I mean I can not run 2600 c8 on the MVF while the Asrock Z77 OC Formula can.
Not even a single iteration passes on the Asus, yet Asrock with less volts passes with ease, not to mention you can pull in TWCL much tighter.
these screens are from Billy
http://i.imgur.com/Sys6z0I.jpg
http://i.imgur.com/QlD94Yo.jpg
My Flares and 2400 c8 fail 32M as well at 2600+ so it is not the dimms as all three kits can run 2600+ on other vendors boards, it must be some setting / timing which I am overlooking.
And this could be the difference as I have not tested a bunch of high mhz memory, but the few kits I have tested all seemed to go very high on ROG boards and clock well, while the Asrock seems to do better with PSC/BBSE for me at least.
Thanks for all the replies guys:up:
@l0ud..Try up vccsa to 1.25 also play with Dram CP, TX/RX slews and MCH ds
Finally had some time to play.3570K + good old sammsungs.
Alex-Ro - 6:19:172 - 50 x 100 - 1400mhz - 9-12-12-23 1T - Asus M5G - 2x4 Corsair Platinum 2666C10 Samsung - 1.9V ; CPU -air /RAM air@23 ambient.
http://img24.imageshack.us/img24/668/snaphsot0002.png
Well done with the 3570K mate :clap:
Thanks,took your advice and scored with 2400 wazza,altough i still miss 0.4s from your score,your run with 3570K was awesome... :D
FINALLY i hit a home run and managed to beat my previous scores with little 3570K :D
Alex-Ro - 6:18:860 - 50 x 100 - 1400mhz - 9-12-12-23 1T - Asus M5G - 2x4 Corsair Platinum 2666C10 Samsung - 1.9V ; CPU -air /RAM air@23 ambient.
Hope someone updates the first page :)
http://img6.imageshack.us/img6/1768/snaphsot0005y.png
He he...... :D
3570K racing...........
C-N - 6:18:859 - 50x100 - 1400MHz - 9-12-12-20-110-1 - Gigabyte UD5H - Corsair Platinum 2666 CL10 2x4GB - 1.88v - Stock Air
http://i106.photobucket.com/albums/m...2126M18859.png
Alex-Ro - 6:18:766 - 50 x 100 - 1400mhz - 9-12-12-23 1T - Asus M5G - 2x4 Corsair Platinum 2666C10 Samsung - 1.92V ; CPU -air /RAM air@15 ambient.
I've nailed it :D Opening the window could make that trwdr=4,unfortunately i can't run it's brother trwsr 4 only at 5 degrees ambient,but....
It's interesting to see we have different approaches,my run starts faster and i have a better final loop,your starts slower but you have very fast middle loops.
http://img834.imageshack.us/img834/3...aphsot0007.png
Awesome mate :up:
looks to me like I better get some beers on the way home from work & have another go.....
Awesome runs Alex and C-N:up:
Go 4X4 and .17s:)
Dumo this is the first gen v1.0 UD5H with the gap between the RAM slots & blue caps it wont even boot 2800 with 4 dims. The UD5H with no gap between the RAM slots v1.0 & 1.1 & the purple caps which I also have will boot 2800 4 x 4 but I cant get it stable or fast.
2 beers down :D
C-N - 6:18:750 - 50x100 - 1400MHz - 9-12-12-20-110-1 - Gigabyte UD5H - Corsair Platinum 2666 CL10 2x4GB - 1.90v - Stock Air
http://i106.photobucket.com/albums/m...2126M18750.png
Awesome run C-N,will try following days,seems C-N tactics with beer is becoming a trademark :D