Yet another UTT Death

[jiff]

Well Im now absolutly convinced this issue is NOTthe Board.
I have had a VX die on a DFI 3 board. Now 1 my 3 week old xp4000 redline has passed away.

Never biosed flashed with these sticks.
Never cold booted computer other than the first boot up.
Never ran without Xflow Fan (stacker)
Never ran without 90 mm Fan over the top
Never ran more than 3.4 volts
Never removed ram
Never used 4v jp17, only jumper trick
Woke up this morning and screen wasnt waking up, computer had frozen. rebooted, nothing!!!! removed the offending stick and here I am typing.

I for one have had enough of UTT looks like I will go back to Ballistix, good mix between UTT and TCCD

[Jaco]

just let the 'bad' stick rest for a couple of days.

then try it again , it might work again. I have seen posts here of bad UTT sticks come alive again.

[teko]

Why not go for BH5, or mushkin redline? (dunno if they are UTT thoug)..

[jiff]

Why not go for BH5, or mushkin redline? (dunno if they are UTT thoug)..

look at sig

[Kabouter]

Why not go for BH5, or mushkin redline? (dunno if they are UTT thoug)..

redline is what died on him, and yes Redline does use UTT

[HousERaT]

I'm quite sure Muskin will replace it free of charge since you've run the memory below 3.5v. People do get bad sticks...... don't give up just yet. The Redline stuff is pretty durable should you get a non-faulty stick.

[ArcTan]

are you still using the 5v rail for vdimm?

[13oost]

may be warrentied up to 3.5 +- 5% but its still hard on the memory, 3.4v can kill it, but yea... let it sit and there is a good chance for it to come back to life

[jiff]

are you still using the 5v rail for vdimm?

Forgot that

I have Never Ran 4v j17, I used bigtoe jumper trick from day one

[jiff]

The Redline stuff is pretty durable should you get a non-faulty stick.

Tis rather early to be saying that, if you look at the mushkin threads there is more and more ppl saying they have dieing sticks. Remember the VX took awhile before we started seeing deaths.

[jiff]

may be warrentied up to 3.5 +- 5% but its still hard on the memory, 3.4v can kill it, but yea... let it sit and there is a good chance for it to come back to life

Well I ran 3.6 through my old BH5 24/7 for years, wish I never soldem

[teko]

look at sig

My bad But then why not BH5?

[13oost]

yea seems that bh-5 hasnt had these problems, but exactly how many bh-5 deaths have there been with these boards, on the "new" and old bh-5

[jiff]

yea seems that bh-5 hasnt had these problems, but exactly how many bh-5 deaths have there been with these boards, on the "new" and old bh-5

Well theres no way that this death or my VX can be the board,
VX on DFI 3
I didnt use 4v jp17 on this Ram, I never cold booted, this ram died failed while PC was on.
I will wait and see how many deaths we see using jumper trick, I bet we see just as many. Ppl are killing UTT with boosters on other boards.

[13oost]

some/alot of the reason that they kill the UTT on other boards with a booster is b/c they dont calibrate the booster before they use it, so theyre craming way more voltage than they think they are...

j/w how many "deaths" have occured on bh-5

[ben805]

at 3.4v these memory still can be smoking hot, I was running 265 @3.5v the other night and I have a Thermaltake 4800rpm (75cfm!) fan blasting on my redline at full speed but after an hour of 3dmark loop the heatspreader were still kinda hot..especially dimm 4 right next to the cpu, make sure you have sufficient cooling on these ram but if they still died atleast we would know heat didn't causes it.

[jiff]

at 3.4v these memory still can be smoking hot, I was running 265 @3.5v the other night and I have a Thermaltake 4800rpm (75cfm!) fan blasting on my redline at full speed but after an hour of 3dmark loop the heatspreader were still kinda hot..especially dimm 4 right next to the cpu, make sure you have sufficient cooling on these ram but if they still died atleast we would know heat didn't causes it.

Well if you have ever seen/used a stacker with xflow flan on full blast and a 90mm Fan sitting on top you would know heat was never an issue

I have indeed ran S&M memory test, touched heat speaders which were cool after 4 hours. The machine was on idle at the time of failure.

Waiting for mushkin to come back to me on this, I would prefer a refund I have no confidence in UTT chips, my guess is UTT products wont be around that much longer.

[wtz54321]

Please check my thread about UTT chip ... http://www.xtremesystems.org/forums/...d=1#post912487

It's quite possible that not all the UTT chip will fail that quickly , but apparently those
quickly failure chip are not sorting out before they put onto the module ...

And what we see now is a typical UTT module failure syndrome , that's why
the UTT chip is cheap ... All the failure case will be treated as RMA issue ...

[drunkenmaster]

utt describes not one chip but several. thats the first thing to mention, you have untested CH-5 mem afaik, thats whats in vx and redline. UTT is not a product, old bh-5 was supposedly rated to 3.6v, the chips, not the pcb's or resistors or anything else. the chips in some circumstances were rated like that, i never heard ch-5, on a smaller die process, was rated as such. its completely up to the companies to decide to charge more, make a profit even with any multiple rma's they have to deal with. you've still bought a product that will run as advertised, and they will replace it, that hasn't changed. if rma rates get too high they'll change price or stop offering it.

very much more than likely the reason vx and redline are on ch and not bh dies is this, the idea behind these chips are cheap as possibly with decent yeilds, on same sized waifers you'll get far more chips on a smaller process.

[wtz54321]

utt describes not one chip but several. thats the first thing to mention, you have untested CH-5 mem afaik, thats whats in vx and redline. UTT is not a product, old bh-5 was supposedly rated to 3.6v, the chips, not the pcb's or resistors or anything else. the chips in some circumstances were rated like that, i never heard ch-5, on a smaller die process, was rated as such. its completely up to the companies to decide to charge more, make a profit even with any multiple rma's they have to deal with. you've still bought a product that will run as advertised, and they will replace it, that hasn't changed. if rma rates get too high they'll change price or stop offering it.

very much more than likely the reason vx and redline are on ch and not bh dies is this, the idea behind these chips are cheap as possibly with decent yeilds, on same sized waifers you'll get far more chips on a smaller process.

If the root cause of the whole thing is UTT , not what kind of UTT ... That explain the situation we see now perfectly ...

[[XC] moddolicous]

did u have a fan over them???

[uwackme]

The chips in question are UTT-CH..... new CH production using some changes from Infineon that yeilded a VAST improvement over OLD CH5 chips.

The "UTT" part refers to noone bothers speed sorting the chips. The BH or CH refers to what the chip actually IS. There is a world of difference between UTT-CH and UTT-BH even though both are "winbond" chips.

The new UTT-BH chips are better than the old BH, and the new UTT-CH chips are even better'r

But we have seen alot of UTT-CH dying.... partly because they need high Vdimm to shine, and get it, and croak. Partly because the are UTT... un sorted and the weak chips dont get ferrett'd out of the bunch before use.

One hopes something like 4000VX and Redline 4000 would get the extra effort of UN-UTT'ing the chips before use. After all, what exactly are oyu paying EXTRA for if not sorting out the best of the best in chips before hand.

But we apparently wont have UTT-CH to kick around anymore as its supply has dried up.

The 8Gigs of it I fielded are all working flawlessly....except my 2 sticks with "high voltage" lockup problem. Im still hoping its the SPD rom that is the actual culprit. The other sticks are happyily running on 3.25-3.45Vdimm systems from NF7-S to DFI Infinity to DFI NF4 SLI-DR.

As for CH, they have the same ABS MAX rating in the data sheet as the BH chips... 3.6V.

[uscfan]

Put your UTT in the freezer, im not kidding. Leave it in there for a few weeks and it should be back to normal.

Ballistix seems pretty sweet though. 285 at near UTT timings, sounds good to me.

How have BH6 been holding up with high voltages? Friend of mine will loan me his Mushkin Black BH6 and he says it will do 260 no problem. Any reports of those dying?

[13oost]

bh is on a 150nm process and the ch is a 130 if i remember correctly.... smaller process pretty much equals less tolerance to higher voltages and heat

ive also hear that the CH chips are thinner and circuits are closer to the surface, but i just remember seeing it mentioned... im not swearing on it

why dont they try putting CH on a larger process like on the BH? it might not be CH then, but if its the same construction then it should preform better and have better tollerances

[jiff]

Put your UTT in the freezer, im not kidding. Leave it in there for a few weeks and it should be back to normal.

Ballistix seems pretty sweet though. 285 at near UTT timings, sounds good to me.

How have BH6 been holding up with high voltages? Friend of mine will loan me his Mushkin Black BH6 and he says it will do 260 no problem. Any reports of those dying?

I would do if it wasnt under warranty, mushkin have just ok a RMA with faulty stick. I guess I will stay with the Redline until it fails then RMA AGAIN and AGAIN