B-Die Binning

[Sparks.nl]

I have got a new (jan 2016) G.Skill 3600 C16 kit.

It does 3866MHz C12-12-12-28 with 1.72v (bios, Asrock overvolt) and XTU (Win 8.1) as a kit.

 

But... It can not boot 1T for 3866MHz

 

I have 2 questions:

 

1. Does it hit on performance verry much? (it seems to affect it)

I did a quick test on 5GHz and got 1768 XTU where I do 1785 on these settings with the other kit)

 

2. Is this common with other kits? (only 2T boot)

 

3600MHz C12 does work with 1T.

[websmile]

Yes, it can happen on some kits but it can also happen that this kit does 1T on another board and IMC. Note that some kits that I saw did 1t after I had initial problems, but they needed much more volts than for 2T

 

P.S. It hits performance, but iirc, it is maybe 1s for spi32m and they are still useful mems

[Sparks.nl]

Thank you for your quick reply.

I noticed something.

 

They do 1T on the same voltages if I use just 1 stick in stead of 2.

How can this be?

 

Is that imc?

[websmile]

IMC or compability, third option is mems cannot take the pressure and bandwidth that dual channel puts on top. But it can also be a weak channel on the board, have you tested both slots seperately?

[Sparks.nl]

I tested in slot 4 only.

 

I think I will try to give a little more volts on vccio and vccsa.

those settings are still set for my other set.

(Which does do 1T)

[phobosq]

Anyone tried 2x8 RipjawsV 3600C17? I took a kit hoping to find B-die that flies. I should have it next Tuesday.

[websmile]

Yes, two kits, worst B-die ever, non could do 3733 12-12

[phobosq]

Hope mine will be better then.

[Sparks.nl]

Yes, two kits, worst B-die ever, non could do 3733 12-12

 

Not even 2T?

[Guest george.kokovinis]

Yes, two kits, worst B-die ever, non could do 3733 12-12

 

 

You should have kept the kit you sold me Michael.

 

3733 / 11-11-28-1T / trfc 200, extremely tight subs. 1.90Vdimm.:ws:

[websmile]

Nope. I know someone wanted to bin OEM and not a single stick out of 20 did 3600 c12

 

The kit is in good hands, I don´t bench at the moment anyway so it is good to see these in use

Edited March 2, 2016 by websmile

[phobosq]

Wasn't that Sam?

[sabishiihito]

Fresh from NewEgg, can't wait to see how they do.

[barti27]

if someone knows they will be spun Memory

 

F4-3733C17D-16GTZA

[TaPaKaH]

found relatively decent kit

 

 

everything is retail/bought with own money, brandnames covered to stop free exposure

[Guest Bullant]

Look good memory Sam

[Splave]

Is that a Yukon Gold?

[marco.is.not.80]

Wow, that's really impressive. 12-11-11 at over 4000 - I can only dream.

 

What I'm curious about is why no one has tested these - not just on this thread but it seems like they aren't getting any attention anywhere:

 

http://www.newegg.com/Product/Product.aspx?Item=N82E16820232206

 

I've had mine for a while and just started playing with them the last few days and they are B-Die for sure (using Typhoon to pull the info - haven't actually cracked them open) but I just figured that a two stick set of high density memory @ low latency would've been something everyone would be jumping on.

 

Please forgive my ignorance if I'm missing something and obviously I'm the little boy at the table compared to the rest of you and maybe at the very least you all get a chuckle :-)

 

- Marco

[websmile]

Tested the 3400c16, good mems, but this is real hard for Z170 and I do not have 2011-3 - in dual channel these need overproportional voltage compared to single stick tests, and on Hero 8 at quad configuration I opted at out 3466 c10 at around 2v, 3466 c11 was ~1,8v at 4x16gb 32m.

http://ocxtreme.net/viewtopic.php?f=8&t=34

Edited April 25, 2016 by websmile

[marco.is.not.80]

Thanks, websmile - that's definitely helpful for me and not sure why I never found your writeup via Google but much appreciated :-)

[l0ud_sil3nc3]

found relatively decent kit

 

 

everything is retail/bought with own money, brandnames covered to stop free exposure

 

 

4.5/4.5 no waza?

 

 

If memory on air, this is the best B die on earth

[Guest Bullant]

I've only bought two kits of retail and still mainly use the first kit of retail I bought just wondering if many people have tried ln2 memory's much on different kits,I've tried cold on my memory's quite a few times and know the temps require for 32m,xtu and max freq all are different temps.On my kit 32m with waza tight for eff is the first bench that looses stability the quickest from memory's being to cold,then xtu and max freq

 

Has anyone seen anything different from kit to kit

Edited April 25, 2016 by Bullant

[rivaldokfc]

hi Bullant，did you mean 32m with waza is easier than xtu test ?

 

 

从我的 iPhone å‘é€ï¼Œä½¿ç”¨ Tapatalk

[Guest Bullant]

hi Bullant，did you mean 32m with waza is easier than xtu test ?

 

 

从我的 iPhone å‘é€ï¼Œä½¿ç”¨ Tapatalk

 

Hey mate,no 32m with waza is the hardest for me on air then xtu then max freq,also same on ln2 memory's 32m doesn't like cold memory's as much as xtu.Soon as temps are to cold 32m will instant fail even if same timings pass on air.

 

Like a perfect memory fir me on ddr4 would be b due timings with AFR cold scaling,just wondering cause I've only tested couple kits that maybe there are b die out there that love cold 32m little more then the kit I have,would be nice to see 32m cold 11-11-11 4133 but seems the memory can boot it on cold but all stability is lost on Pi 32 when you go past temp

 

Seems to be the norm or typical Samsung thing from what I've seen

Edited April 26, 2016 by Bullant

[rivaldokfc]

thanks Bullant. I had a kit that do 4000-12-11-11-2t 32m on atx ocf. but after 2-3 times of running, the kit failed everytime....do you have any expeeience on this situation?

 

 

从我的 iPhone å‘é€ï¼Œä½¿ç”¨ Tapatalk